Aberrant splicing in human cancer has a large-scale functional impact on transcription factors

Aberrant splicing in human cancer has a large-scale functional impact on transcription factors | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Aberrant splicing in human cancer has a large-scale functional impact on transcription factors Khalique Newaz, Olga Tsoy, Jan Baumbach This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7799323/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Cancer shows aberrant alternative splicing (AS), which could functionally perturb proteins required for normal cellular behavior. Transcription factors (TFs) are proteins that regulate gene transcription. Some evidence about AS-driven perturbation of TFs in cancer has been documented. A comprehensive systematic analysis of how cancer-specific AS could affect functions of TFs is missing. Such an analysis could reveal the molecular mechanisms of cancer progression due to transcription misregulations, and thus could help identify therapeutic targets. Here, we systematically analyzed potential functional perturbations of TFs due to AS across 15 cancer types. By analyzing AS patterns, we identified 2 118 perturbed AS events (i.e., events showing significant AS pattern differences between normal and paired cancer samples) that affect 718 TFs across 14 cancer types. In 189 TFs, the perturbed AS events affected known functional domains of TFs. Indirect evidence for the functional impact of cancer-specific AS on TFs were also found: First, by relating AS patterns of the perturbed AS events with the DNA-binding and regulatory activity of TFs. Second, by using cancer dependency data to explore whether the affected TFs are essential for cancer cell line proliferation. Our findings show a large-scale functional perturbation of TFs due to cancer-specific AS. Biological sciences/Cancer Biological sciences/Molecular biology/Transcriptomics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1. Introduction Alternative splicing (AS) results in multiple different messenger ribonucleic acids (mRNAs) originating from a single gene[ 1 ]. AS is associated with normal cellular processes, including tissue development and identity[ 2 ]. Normal AS patterns are perturbed in diseases[ 3 ], including in cancer[ 4 ]. Large-scale RNA sequencing (RNAseq)-based evidence for aberrant AS in cancer has been collected. For example, the cancer genome atlas (TCGA) SpliceSeq database contains RNAseq-based quantification of AS events across 33 cancer types[ 5 ]. TCGA SpliceSeq has been used to explore cancer-specific AS[ 6 ]. For example, Yongsheg et al.[ 7 ] investigated relationships between somatic mutations in cancer and cancer-specific AS patterns, and found that different mutations in the same gene showed distinct AS perturbations. Junyi et al.[ 8 ] exploited the hypothesis that RNA-binding proteins (RBPs) regulate cancer-specific AS patterns and proposed a computational pipeline to identify RBPs that are cancer-specific AS regulators. Similarly, Cheng et al.[ 9 ] investigated whether abnormal AS of splicing factors could explain cancer-specific AS patterns. Aberrant AS in cancer has also been shown to impact an important group of proteins called transcription factors (TFs)[ 10 ]. TFs bind DNA to regulate transcription of genes[ 11 ]. Cancer-specific dysregulation of TFs perturbs normal transcriptional programs, which could lead to a sustained tumor state[ 12 ]. Hence, TFs have been considered potential cancer therapeutic targets[ 13 ]. TF dysregulations could be due to cancer-specific signaling cascades that perturb TF expression or changes in the corresponding gene structure via chromosomal translocations/point mutations[ 14 ]. Some evidence of cancer-specific AS effects on TFs has also been documented[ 10 , 15 ]. For example, AS-induced alternative protein isoforms of T Cell Factor 4 ( TCF4 ) have been implicated in proliferations of non-small cell lung cancer cells[ 16 ] and hepatocellular carcinoma[ 17 ]. Other examples are reviewed elsewhere[ 10 ]. Despite existing evidence, a comprehensive systematic analysis of how cancer-specific AS affects TFs has been missing. Such an analysis could reveal how TF-affecting AS events drive cancer progression, which could then be used to identify therapeutic targets and strategize drug design[ 18 ]. Here, TCGA SpliceSeq was used to systematically explore the effect of cancer-specific AS events across all known 1 639 human TFs. 15 cancer types that had AS data for enough paired (cancer and healthy) patient samples were used. For each cancer-type, AS events showing statistically significant differences between paired cancer and normal samples were considered as perturbed. 2 118 TF-affecting perturbed AS events covering 718 TFs across 14 cancer types were found. Comparisons across cancer types revealed several unique cancer type-specific, as well as common, TF-affecting perturbed AS events. In 189 of the 718 (~ 26%) TFs, at least one perturbed AS event affects TF functional domains. This provides direct evidence that cancer-specific AS events have a functional impact on TFs. Additionally, we found indirect evidence of the functional impact of cancer-specific AS events on TFs in two ways. First, we found that the expression patterns of > 100 AS events covering > 70 of the 718 TFs have significant correlations with the existing sample-matched TF footprinting data[ 19 ], implying a possible dependence of the DNA binding and gene regulatory activities of the affected TFs on the corresponding perturbed AS events. Second, utilizing existing cancer dependency data[ 20 ], we found that many (39) TFs affected by perturbed AS events are also essential for cancer progression in cancer cell lines. Overall, our findings highlight a large-scale functional impact of cancer-specific AS on TFs. The collected evidence could guide future experiments to understand cancer molecular mechanisms or identify therapeutic targets. 2. Data and methods 2.1 Alternative splicing data From the TCGA database[ 21 ] (accessed Feb. 2025), 15 cancer types were selected, where each cancer type had at least 10 patients with primary tumor (cancer) and patient-matched control (healthy) samples (Table 1 ). For these paired samples, the corresponding AS data were downloaded from TCGA SpliceSeq[ 5 ]. This data contains information about seven types of AS events: Alternative Acceptor (AA), Alternative Donor (AD), Alternative Promoter (AP), Alternative Terminator (AT), Exon Skip (ES), Mutually Exclusive Exons (ME), and Retained Intron (RI). To quantify AS events in a TCGA sample, TCGA SpliceSeq provides Percent Spliced In (PSI) values for AS events covered by at least eight RNA-seq reads; for the rest of the AS events, a “null” value is provided. For each cancer type, we only considered events with PSI values for at least 10 paired samples. Table 1 Patient counts across 15 cancer types in TCGA SpliceSeq. Cancer type TCGA cancer code # of patients with paired samples # of cancer samples # of AS events with values for at least 10 paired samples Bladder urothelial carcinoma BLCA 18 406 42 343 Breast invasive carcinoma BRCA 96 1094 58 913 Colon adenocarcinoma COAD 39 458 48 843 Esophageal carcinoma ESCA 13 180 44 778 Head and neck squamous cell carcinoma HNSC 42 501 51 442 Kidney chromophobe KICH 25 66 50 059 Kidney renal clear cell carcinoma KIRC 71 533 57 286 Kidney renal papillary cell carcinoma KIRP 32 290 52 337 Liver hepatocellular carcinoma LIHC 50 371 43 753 Lung adenocarcinoma LUAD 51 514 51 436 Lung squamous cell carcinoma LUSC 48 501 57 040 Prostate adenocarcinoma PRAD 51 497 55 799 Stomach adenocarcinoma STAD 33 415 54 446 Thyroid carcinoma THCA 57 505 56 824 Uterine corpus endometrial carcinoma UCEC 23 545 36 585 For each cancer type, significantly perturbed AS events were detected as follows. For an AS event, only the patients with PSI values for both normal and cancer samples were detected. Considering these patients, a two-sided paired Wilcoxon Signed Rank test was performed to compare PSI values from normal samples and the paired cancer samples using the “ wilcox.test ” function in R to obtain a p-value. The p-values across all AS events were corrected using the Benjamini Hochberg procedure to obtain the corresponding q-values. An AS event was considered as significantly perturbed if and only if (1) its q-value was 5%. 2.2 Transcription factors 2.2.1 Curation and mapping of genomic coordinates All 1 639 curated human TFs were obtained from Supplementary Table S1 of the Lambert et al.[ 11 ] study. For these TFs, canonical protein sequences from UniProt were collected. Henceforth, by the term “UniProt protein sequence” we mean the “canonical UniProt protein sequence”. UniProt IDs and Ensembl transcripts were matched using UniProt. Because one TF can have multiple Ensembl transcripts, one transcript per TF was selected as follows. Given a TF, its UniProt sequence was pairwise globally aligned with each of the corresponding Ensembl transcripts. Protein sequences of the transcripts were constructed using the release 113 of Genome Reference Consortium Human Build 37 (GRCh37) Ensembl GTF file “Homo_sapiens.GRCh37.dna.primary_assembly.fa.gz”. Then, the transcript that matched the TF protein sequence with 100% identity and no gaps was selected. For the 1 469 TFs with matched transcripts as per the above criterion, genomic coordinates were mapped onto their protein sequences using GRCh37. Note that GRCh37 was used instead of GRCh38 because TCGA SpliceSeq used GRCh37 to map genomic coordinates of the AS events. Genomic coordinates of AS events provided by TCGA SpliceSeq were then used to mark parts of the TF protein sequences affected by AS events. Specifically, for each of the six types of AS events, i.e., AA, AD, AP, AT, ES, and ME, the following was done. Note that RI AS events can not be mapped to TFs as they do not overlap with annotated protein-coding regions defined by GRCh37. For AA, AD, ES, or ME, all sequence positions of a TF whose start and stop genomic coordinates lie within the start and stop coordinates of the affected exons of an AS event were marked as affected. For AP, all sequence positions of a TF whose start genomic coordinates lie before the start position of the affected exon were marked as affected. For AT, all sequence positions of a TF whose stop genomic coordinates lie beyond the stop position of the affected exon were marked as affected. 2.2.2 Mapping of DNA-binding and effector domains UniProt collects DNA-binding domain (DBD) information of proteins using PROSITE[ 22 ], Pfam[ 23 ], or SMART[ 24 ]. This DBD information was used to annotate which sequence positions of TFs belong to a DBD region. Specifically, given the information about a TF in UniProt, if present, the following feature table entries were considered as a DBD: (1) entries “DNA_BIND” or “ZN_FING” or (2) entry “DOMAIN” with domain names 'bHLH', 'bZIP', or 'MADS-box'. Of the 1 469 TFs with a matching Ensembl transcript (Section 2.2.1 ), 1 347 had at least one sequence position belonging to a DBD. TF effector domain (ED) information was taken from Supplementary Tables S2 and S3 of the Soto et al.[ 25 ] study. This study looked for ED evidence across 1 639 human TFs in the literature and manually curated the largest collections of 924 EDs across 594 TFs. For these TFs information about the UniProt sequence (the corresponding isoform) positions of EDs are provided. This ED information was used to annotate which UniProt sequence positions of TFs belong to an ED region. Of the 1 469 TFs (Section 2.2.1 ), 460 had at least one sequence position belonging to an ED region, covering 730 EDs. 2.3 tSNE application Across all 15 cancer types, 2 278 TF AS events were identified that were present in each cancer type. Among these, 1 011 TF AS events were present as perturbed TF splicing events (PTSEs) in at least one cancer type. PSI patterns of these 1 011 AS events for all 6 876 cancer samples across the 15 cancer types were used as input to the “ Rtsne ” package of R to obtain dimensionally reduced representations of the cancer samples. 2.4 Hypergeometric test of DBD perturbations Among the 1 347 TFs with DBD annotation (Section 2.2.2 ), 1 464 occurrences of DBDs were found, which were taken as the background size. Among all PTSEs across all cancer types, 146 unique DBDs were present, which were taken as the sample size. For each of the 27 DBD-types perturbed in at least one PTSE, their counts in the sample and the background were used to compute the corresponding over-representations via the hypergeometric test to obtain 27 p-values. Similarly, for each of the 27 DBD-types, their counts in the sample and the background were used to compute the corresponding under-representations via the hypergeometric test to obtain 27 p-values. The 54 hypergeometric p-values were corrected using the Benjamini Hochberg procedure to obtain the corresponding q-values. 2.5 Alternative splicing events significantly associated with patient survival Associations between AS events from TCGA SpliceSeq and cancer patient survival were obtained from the OncoSplicing database (“SpliceSeq_info_Survival.csv.gz”) developed in Zhang et al.[ 26 ]. For a cancer type, OncoSplicing contains AS events that correlate with clinical outcomes according to the Cox proportional hazards regression. All AS events with survival association p-values of less than 0.05 (reported in the column named ‘Pvalue_HR’ of “SpliceSeq_info_Survival.csv.gz”) were considered as significant and selected for our study. 2.6 Correlation of PSI values of PTSEs and TF footprinting TF footprinting data were taken from Supplementary Table S6 of Corces et al.[ 19 ], which covers 410 TCGA cancer samples and 843 TFs across 23 cancer types. Corces et al. provides footprinting depth and flanking accessibility values for 1 781 TF-DNA motif pairs, with multiple DNA motifs for a given TF. For a cancer type, a correlation between PSI values of a PTSE and footprinting values of a TF-DNA motif pair was measured if and only if there were at least 10 cancer samples for which the PTSE had a non-null PSI value in TCGA SpliceSeq and footprinting data in Corces et al. At least 10 cancer samples were considered to retain power in the correlation analysis. 769 PTSEs across 12 cancer types fulfilled this criterion (Table 2 ). Table 2 Overlap between TCGA SpliceSeq and Corces et al. 19 . TCGA cancer code # of cancer samples # of PTSEs with PSI values for at least 10 cancer samples # of TFs within PTSEs with PSI values for at least 10 cancer samples BLCA 10 42 28 BRCA 74 296 132 COAD 38 176 88 KIRC 16 245 116 KIRP 33 150 84 LIHC 17 93 37 LUAD 20 198 93 LUSC 16 299 135 PRAD 26 124 67 STAD 21 96 56 THCA 14 144 78 UCEC 13 85 52 Given a cancer type, for each PTSE selected as per the above criterion, we measured the correlation between its PSI values and footprinting depth of the corresponding TF-DNA motif pairs. Over all 12 cancer types this resulted in 3 676 correlation pairs. Similarly, we obtained 3 676 correlation pairs corresponding to flanking accessibility. We calculated the empirical p-value based on the data randomization. For each of the 3 676*2 = 7 352 pairs, 100 random correlations were computed by randomly shuffling the PSI values and the footprinting data. The fraction of random correlations higher than the real correlation was taken as the empirical p-value. The resulting 7 352 p-values were corrected using the Benjamini Hochberg procedure to obtain the corresponding q-values. A q-value < 0.05 was considered significant. 2.7 Dependency profiles of TFs associated with PTSEs Dependency profiles of cancer cell lines were downloaded from the depmap[ 20 ] portal (“ https://depmap.org/portal/data_page/?tab=customDownloads ”) , which included all available cell lines and genes. This data cover 12 cancer types considered in our study: BLCA, BRCA, COAD, HNSC, KIRC, LIHC, LUAD, LUSC, PRAD, STAD, THCA, and UCEC. For these 12 cancer types, cell lines corresponding the following terminologies (in the order of the cancer types mentioned above) from the “lineage_3” column of the data was used: “Bladder Urothelial”, “Breast Invasive Carcinoma”, “Colorectal Adenocarcinoma”, “Head and Neck Squamous Cell Carcinoma”, “Renal Clear Cell Carcinoma”, “Hepatocellular Carcinoma”, “Lung Adenocarcinoma”, “Lung Squamous Cell Carcinoma”, “Prostate Adenocarcinoma”, “Stomach Adenocarcinoma”, “Thyroid Cancer”, “Endometrial Carcinoma”. The PTSEs across all of the above 12 cancer types cover 688 TFs. Of the 688 TFs, for 669 TFs across the 12 cancer types, there is dependency data with respect to at least one cancer cell line, totaling 2 505 TF-cancer type combinations. 3. Results and discussion 3.1 Perturbed TF splicing events in individual cancer types Given a cancer type, all AS events showing statistically significant (q-value 5% between the means of PSI values from normal and cancer samples) PSI differences were considered as perturbed AS events (Section 2.1 ). From these, all AS events corresponding to TFs (Section 2.2 ) were considered as perturbed TF splicing events (PTSEs). Over all cancer types, 718 TFs were affected by at least one PTSE. Many PTSEs were found for each cancer type, except for ESCA that showed no perturbed AS event and thus no PTSE. No perturbed AS event for ESCA could be due to the low number of samples (13; see Table 1 ) that do not provide enough statistical power to detect significant differences between cancer and normal samples with the current q-value cutoff. The number of PTSEs showed large variations across cancer types, with the highest number (750) of PTSEs for LUSC (Fig. 2 A). On average over all cancer types, ~ 1.8 PTSEs came from every TF, implying that no TF dominates PTSEs. PTSE numbers within cancer types do not seem to correlate with the corresponding numbers of paired normal and cancer samples (Table 1 ). For example, LIHC, LUAD, LUSC, and PRAD, had similar numbers (50, 51, 48, and 51 respectively) of paired samples, but their PTSE numbers varied substantially: 247, 461, 750, and 310, respectively. Overall, these results indicate that not only are there significant TF AS differences between normal and cancer samples within a cancer type, but these differences also vary considerably across cancer types. For many PTSEs within a cancer type, their PSI values of cancer samples were higher/lower than their PSI values of the paired normal samples for a majority of patients (Fig. 2 B). For example, the PSI values for the AS event HMGA2_22880_AT (here, “HMGA2” is the gene symbol, “22880” is the AS event number provided by the TCGASpliceSeq database, and “AT” is the type of AS event) were higher in cancer than normal samples for 52 out of the 57 paired samples of THCA with a mean ΔPSI of around 0.44 (Supplementary Table S1 ). Here, ΔPSI was defined as the difference between the PSI value of a cancer sample and the paired normal sample, with mean ΔPSI denoting the average ΔPSI over all paired samples of a given cancer type. The above results imply cancer type relevance of PTSEs. To further check whether this is the case, we used t-distributed Stochastic Neighbor Embedding (tSNE) to visualize cancer samples of the entire cohort of the considered cancer types, based on PSI patterns of all 1 011 AS events that were present in each of the 15 cancer types and were present as PTSEs in at least one cancer type (Section 2.3 ). We found that cancer type-related samples were generally clustered together (Fig. 2 C), implying cancer tissue specificity of the PSI patterns of PTSEs. Similar results have been shown using PSI patterns of all AS events in cancer[ 8 , 9 ]. Interestingly, although we found large variations in PTSEs with respect to their numbers and PSI patterns across cancer types, the proportions of different types of AS events within PTSEs were similar across cancer types. AP, AT, and ES dominate PTSEs (Fig. 2 D), which has also been found in existing studies using all AS events[ 8 , 9 ]. 3.2 Common and unique PTSEs of individual cancer types Of the total 2 118 PTSEs, a majority (~ 61%, i.e., 1 302 PTSEs) occurred in more than one cancer type (Fig. 3 A and Supplementary Table S2 ). 39 PTSEs occurred in at least 10 cancer types (Fig. 3 B). An example PTSE is PHD19_87401_AT that affects the PHD Finger Protein 19 ( PHD19 ) gene. This PTSE was upregulated in cancer samples in 12 cancer types, and corresponds to a known PHF19 isoform (hPCL3s) that is related to cancer progression of multiple cancer types[ 27 – 30 ]. We also found many unique PTSEs (816 in total) for individual cancer types that varied across cancer types, with BRCA having the most (151) and BLCA having the least (6) of them (Fig. 3 C). Interestingly, at least one of the unique PTSE is survival-associated with the corresponding cancer type (Section 2.5 ), but not to any of the other cancer types (Fig. 3 C and Supplementary Table S3 ). For example, PTSE PITX2_70356_AP affects the Pituitary homeobox 2 ( PITX2 ) gene that results in the PITX2 isoform named PITX2D[ 31 ]. PITX2_70356_AP was upregulated in cancer samples compared to normal samples in BLCA with a mean ΔPSI of 0.27, and was found to be associated with disease-free interval of BLCA with a hazard ratio of 0.31 (Supplementary Table S3 ). This means that an increase in the inclusion of this PTSE indicates a decreased risk of BLCA recurrence. PITX2 isoforms have been shown to be crucial for developmental processes, with PITX2D inhibiting the transcriptional activities of two other PITX2 isoforms by physically interacting with them[ 31 ]. This might explain why the increased expression of PITX2_70356_AP , which potentially leads to an increased expression of PITX2D, correlates with a decrease in the risk of BLCA recurrence. These results suggest that unique PTSEs could be relevant to understand the molecular mechanisms of disease progression of individual cancer types. 3.3 PTSEs perturb DNA-binding domains DBDs are important functional components of TFs, which enable TFs to recognize and bind the specific DNA sequences related to their target genes[ 11 ]. AS events affecting DBDs can impact the DNA-binding ability of the resulting TF isoforms. To capture this, we evaluated whether PTSEs affect DBD regions of TFs (Section 2.2 ). Over all cancer types, DBD regions of 135 TFs (i.e., ~ 10% of all 1 347 TFs with a DBD information; Section 2.2 ) were perturbed by at least one PTSE (Supplementary Table S4 ). Similar to the numbers of all PTSEs (Fig. 3 A), numbers of DBD perturbing PTSEs show large variations across cancer types, with the highest number (87) for LUSC and the lowest number (8) for BLCA (Fig. 4 A). Many of such PTSEs affect large portions of DBD regions within protein sequences of TFs (Fig. 4 B), and thus could have a strong functional impact. An interesting example is the PTSE ELF5_14951_AT , which perturbs the Erythroblast Transformation Specific (ETS) domain of E74-like transcription factor 5 (ELF5). This PTSE has higher PSI values in cancer samples than in normal samples of kidney cancer: KICH, KIRP, and KIRC (Fig. 4 B). ELF5_14951_AT seldom shows large differences among normal vs. cancer samples in other cancer types (Supplementary Table S1 ), implying its potential functional relevance in kidney cancer. Interestingly, ELF5 has been shown to be a lineage-specific TF of principal cells that are key cells involved in kidney development[ 32 ]. Overall, these results indicate that PTSEs could have functional relevance in cancer phenotype development. We next evaluated whether certain DBD types are significantly over/under-represented within PTSEs. Specifically, we quantified the number of DBDs within PTSEs of any cancer type compared to the number of DBDs over all considered TFs, using the hypergeometric test (Section 2.4 ). The proportions of DBDs among all considered TFs and those present in PTSEs were similar, with one significant exception (Fig. 4 C). That is, the Homeobox domain was significantly under-represented (q-value < 10 − 06 ) in PTSEs. Under-representation of the Homeobox domain in cancer was also observed in a recent study that detected 1 350 TF-DNA interactions among 265 TFs and cloned promoters of 108 cancer-related genes[ 33 ]. These results mean that the Homeobox domain is not only less relevant for cancer-related gene regulations in general, but it is also less affected by cancer-related AS events. Within individual cancer types, the DBD types show different perturbation patterns (Fig. 4 D). Some DBD types (i.e., bHLH, bZIP, C2H2-type, IRF tryptophan pentad repeat, PHD-type, and THAP-type) were perturbed in multiple (> 12) cancer types, while others (i.e., A.T hook, CCHHC-type, and Homeobox) were perturbed in selected (< 3) cancer types. These results could reflect the underlying AS-induced shared and unique TF-DNA binding perturbations among cancer types. 3.4 PTSEs perturb effector domains Besides DBDs, effector domains (EDs) also constitute important functional components of TFs that enable activation or repression of the target genes[ 25 ]. Thus, AS events affecting EDs can also impact the functionality of the resulting TF isoforms. To capture this, we evaluated whether PTSEs affect ED regions of TFs (Section 2.2 ). Similar to the numbers of DBD perturbing PTSEs (Fig. 4 A), numbers of ED perturbing PTSEs show large variations across cancer types, with the highest number (51) for KIRC and the lowest number (14) for BLCA (Fig. 5 A and Supplementary Table S5 ). Similar to DBD perturbing PTSEs, many ED perturbing PTSEs affect large portions of ED regions within TF protein sequences (Fig. 4 B). For example, the PTSE TP73_321_AP perturbs the activator domain (AD) of TP73 in LUSC. This PTSE has higher PSI values in cancer samples than in normal samples of LUSC (Fig. 5 B). TP73_321_AP leads to the ΔNp73 isoform of TP73, which is known to be overexpressed in tumors compared to normal conditions[ 34 ] and has been shown to fail in inducing apoptosis and cell cycle arrest[ 35 ]. Over all cancer types, 149 PTSEs perturb ED regions across 84 TFs (i.e., ~ 18% of all 460 TFs with ED information; Section 2.2 ). This proportion is almost twice the proportion of TFs (~ 10%) whose DBD regions were perturbed by PTSEs. This implies that PTSEs more frequently affect the transcriptional activity of TFs than their DNA binding ability. Of all 718 TFs that are perturbed by at least one PTSE, 189 (~ 26%) involve a DBD or an ED, of which 29 TFs show perturbations in both ED and DBD regions. These results imply that AS causes large-scale functional perturbation of TFs in cancer. 3.5 Alternative splicing patterns of PTSEs correlate with TF footprinting To determine potentially AS-driven TF binding/activity patterns, we correlated the PSI values of PTSEs with the TF footprinting of the corresponding TFs in cancer samples of individual cancer types (Section 2.6 ). To do this, the chromatin accessibility data from Corces et al.[ 19 ] was used. Corces et al. used assay for transposase-accessible chromatin using sequencing (ATAC-seq) to determine chromatin occupancy of TFs in TCGA cancer samples, quantified via “footprinting depth” and “flanking accessibility”. Intuitively, footprinting depth measures the inverse of the extent to which the binding of a TF to a chromatin region protects the corresponding TF-binding DNA motif from DNase[ 36 ]. Thus, low footprinting depth means high TF binding, and vice versa. In contrast, flanking accessibility measures the accessibility of the DNA region adjacent to TF-binding DNA motifs, which has been associated with TF activity[ 37 ]. Thus, high flanking accessibility could mean high TF activity, and vice versa. We found 122 PTSEs covering 77 TFs across 11 cancer types (146 unique PTSE-cancer type combinations) that showed significant (q-value < 0.05) correlation with footprinting depth, flanking accessibility, or both (Fig. 6 and Supplementary Table S6 ). Many of these PTSEs showed high mean ΔPSI values. For example, the PTSE PPARG_63413_AP affecting Peroxisome Proliferator-Activated Receptor Gamma (PPARG) in LUSC is significantly negatively correlated with footprinting depth and positively correlated with flanking accessibility. This could imply that, in LUSC, with an increase in the proportion of PPARG_63413_AP there is an increase in the DNA binding as well as the gene regulatory activity of PPARG. Because PPARG_63413_AP shows a mean ΔPSI of -0.42 in LUSC, DNA binding and gene regulatory activities of PPARG is perhaps higher in the normal tissue than in LUSC. Interestingly, previous in vivo[ 38 ] and in vitro[ 39 ] studies have shown that an increase in the PPARG activity slows down the tumor growth rate in non-small-cell lung cancer. Overall, this analysis identified PTSEs within cancer types whose inclusion rates are correlated with the activities of the corresponding TFs. 3.6 PTSEs include TFs showing cancer dependencies Loss-of-function screens in cancer cell lines, e.g., using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)[ 40 ], have identified genes essential for cancer cell line survival. Because many TFs have been associated with the sustained cancer cell state[ 12 ], we evaluated the extent to which the TFs within PTSEs show cancer dependency, using the CRISPR-Cas9 screening data from The Cancer Dependency Map (DepMap, Broad (2025). DepMap Public 25Q2. Dataset. depmap.org)[ 20 ] (Section 2.7 ). Given a gene and a cancer cell line, DepMap provides gene perturbation effects based on Chronos score. A Chronos score of 0 or above indicates that the gene is not essential for the cancer cell proliferation, while a Chronos score of below 0 indicates the gene is essential for the cancer cell line proliferation, with the more negative value indicating more essentiality. We found 669 TFs covered by PTSEs across 12 cancer types that had cancer dependency data for at least one cancer cell line (Fig. 7 A). 646 of the 669 TFs showed a negative Chronos score in cell lines associated with at least one cancer type (Supplementary Table S7 ). 39 of these TFs showed high (Chronos score < -1) dependency in a cancer cell line (Fig. 7 B-M). An interesting example is the PTSE NKX2-1_27289_ES affecting Homeobox protein Nkx-2.1 (NKX2-1) in LUAD (mean ΔPSI of -0.3). NKX2-1 has two known isoforms[ 41 ], where the PTSE NKX2-1_27289_ES leads to a shorter protein isoform. Because the mean ΔPSI value of NKX2-1_27289_ES is negative, the shorter isoform NKX2-1 potentially shows higher prevalence in cancer samples than in normal samples. NKX2-1 is a lineage TF for lung[ 42 ]. Functional differences between the two isoforms have also been shown[ 43 ]. Because the minimum Chronos score of NKX2-1 is -1.4 (Supplementary Table S7 ), indicating its essentiality in cancer proliferation, perhaps it's the higher prevalence of the shorter isoform that determines the cancer dependency of NKX2-1. Overall, this analysis identified TFs that are not only essential for cancer proliferation, but also showed significant differences between normal and paired cancer samples with respect to AS events. This could imply that cancer dependency of these TFs are reliant on the corresponding TF isoforms. 4 Discussion Many AS events showing significant differences between paired normal and cancer samples were found across 14 cancer types, affecting nearly 50% of known TFs with sufficient annotations. Several of these TFs were found to be common in multiple cancer types, and many others were uniquely affected in a cancer type. These results highlight a pancancer significance of aberrant TF splicing in cancer. We found evidence of the impact of cancer-specific AS events on the DNA-binding and gene regulatory (activating/repressing) abilities of TFs. Further evidence for the functional impact of cancer-specific AS events on TFs were found by exploring whether the affected TFs were essential for cancer cell line proliferation. Overall, these findings could be used by future research to understand AS-driven cancer molecular mechanisms, which could guide identifications of novel therapeutic targets. Different algorithms to detect and quantify AS events have been proposed in the literature, e.g., SpliceSeq[ 44 ], Sp1Adder[ 45 ], and Cufflinks[ 46 ]. All of the findings in this work are based on the TCGA SpliceSeq database, which used the SpliceSeq algorithm to quantify AS events. Future work could evaluate whether these findings are robust across different algorithms. Gene regulatory associations based on gene level expressions of TFs have provided interesting insights in cancer[ 47 ]. Our results show that there is a large-scale potential functional impact of cancer-specific AS events on TFs, which could perturb gene regulatory associations. Future inference of gene regulatory associations in cancer could benefit from considering expressions of TFs not just on the gene level but also on the isoform level. Declarations Author contributions KN and JB conceived the study. KN performed all of the work. KN, OT, and JB analyzed the results. JB provided the computational infrastructure. KN, OT, and JB wrote the manuscript. Competing interests The authors claim no competing interest. Data and code availability All data were taken from public databases and the relevant links have been provided within the manuscript. The code to reproduce the work is publicly available at https://doi.org/10.5281/zenodo.17279821. The data extracted from the public databases for reproducing this work is available at https://doi.org/10.5281/zenodo.17250324. Funding This work was supported by the Excellence Strategy of the Federal Government of Germany and the Länder, the Deutsche Forschungsgemeinschaft (grant number 517063424), and the Bundesministerium für Bildung und Forschung (grant numbers 01ZX1908A, 01ZX2208A, and 031L0287B). References Pan Q, Shai O, Lee LJ, Frey BJ, Blencowe BJ. Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet. 2008;40(12):1413–5. Baralle FE, Giudice J. Alternative splicing as a regulator of development and tissue identity. Nat Rev Mol Cell Biol. 2017;18(7):437–51. Scotti MM, Swanson MS. RNA mis-splicing in disease. Nat Rev Genet. 2016;17(1):19–32. Stanley RF, Abdel-Wahab O. Dysregulation and therapeutic targeting of RNA splicing in cancer. Nat Cancer. 2022;3(5):536–46. Ryan M, Wong WC, Brown R, Akbani R, Su X, Broom B, Melott J, Weinstein J. TCGASpliceSeq a compendium of alternative mRNA splicing in cancer. Nucleic Acids Res. 2016;44(D1):D1018–22. Sveen A, Kilpinen S, Ruusulehto A, Lothe RA, Skotheim RI. Aberrant RNA splicing in cancer; expression changes and driver mutations of splicing factor genes. Oncogene. 2016;35(19):2413–27. Li Y, Sahni N, Pancsa R, McGrail DJ, Xu J, Hua X, Coulombe-Huntington J, Ryan M, Tychhon B, Sudhakar D, Hu L, Tyers M, Jiang X, Lin SY, Babu MM, Yi S. Revealing the determinants of widespread alternative splicing perturbation in cancer. Cell Rep. 2017;21(3):798–812. Li J, Pan T, Chen L, Wang Q, Chang Z, Zhou W, Li X, Xu G, Li X, Li Y, Zhang Y. Alternative splicing perturbation landscape identifies RNA binding proteins as potential therapeutic targets in cancer. Mol Ther Nucleic Acids. 2021;24:792–806. Cheng R, Xiao L, Zhou W, Jin X, Xu Z, Xu C, Wang P, Luo M, Wang M, Ma K, Cao H, Huang Y, Lin X, Pang F, Li Y, Jiang Q. A pan-cancer analysis of alternative splicing of splicing factors in 6904 patients. Oncogene. 2021;40(35):5441–50. Belluti S, Rigillo G, Imbriano C. Transcription factors in cancer: When alternative splicing determines opposite cell fates. Cells. 2020;9(3):760. Lambert SA, Jolma A, Campitelli LF, Das PK, Yin Y, Albu M, Chen X, Taipale J, Hughes TR, Weirauch MT. The human transcription factors. Cell. 2018;172(4):650–65. Bradner JE, Hnisz D, Young RA. Transcriptional addiction in cancer. Cell. 2017;168(4):629–43. Bushweller JH. Targeting transcription factors in cancer - from undruggable to reality. Nat Rev Cancer. 2019;19(11):611–24. Darnell JE Jr. Transcription factors as targets for cancer therapy. Nat Rev Cancer. 2002;2(10):740–9. Lambourne L, Mattioli K, Santoso C, Sheynkman G, Inukai S, Kaundal B, Berenson A, Spirohn-Fitzgerald K, Bhattacharjee A, Rothman E, Shrestha S, Laval F, Carroll BS, Plassmeyer SP, Emenecker RJ, Yang Z, Bisht D, Sewell JA, Li G, Prasad A, Phanor S, Lane R, Moyer DC, Hunt T, Balcha D, Gebbia M, Twizere JC, Hao T, Holehouse AS, Frankish A, Riback JA, Salomonis N, Calderwood MA, Hill DE, Sahni N, Vidal M, Bulyk ML, Fuxman Bass JI. Widespread variation in molecular interactions and regulatory properties among transcription factor isoforms. Mol Cell. 2025;85(7):1445–66.e13. Fan YC, Min L, Chen H, Liu YL. Alternative splicing isoform of T cell factor 4K suppresses the proliferation and metastasis of non-small cell lung cancer cells. Genet Mol Res. 2015;14(4):14009–18. Tsedensodnom O, Koga H, Rosenberg SA, Nambotin SB, Carroll JJ, Wands JR, Kim M. Identification of T-cell factor-4 isoforms that contribute to the malignant phenotype of hepatocellular carcinoma cells. Exp Cell Res. 2011;317(7):920–31. Zhang Y, Qian J, Gu C, Yang Y. Alternative splicing and cancer: a systematic review. Signal Transduct Target Ther. 2021;6(1):78. Corces MR, Granja JM, Shams S, Louie BH, Seoane JA, Zhou W, Silva TC, Groeneveld C, Wong CK, Cho SW, Satpathy AT, Mumbach MR, Hoadley KA, Robertson AG, Sheffield NC, Felau I, Castro MAA, Berman BP, Staudt LM, Zenklusen JC, Laird PW, Curtis C, Cancer Genome Atlas Analysis Network, Greenleaf WJ, Chang HY. The chromatin accessibility landscape of primary human cancers. Science. 2018;362(6413):eaav1898. Arafeh R, Shibue T, Dempster JM, Hahn WC, Vazquez F. The present and future of the Cancer Dependency Map. Nat Rev Cancer. 2025;25(1):59–73. Cancer Genome Atlas Research Network, Weinstein JN, Collisson EA, Mills GB, Shaw KRM, Ozenberger BA, Ellrott K, Shmulevich I, Sander C, Stuart JM. The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet. 2013;45(10):1113–20. Sigrist CJA, de Castro E, Cerutti L, Cuche BA, Hulo N, Bridge A, Bougueleret L, Xenarios I. New and continuing developments at PROSITE. Nucleic Acids Res. 2013;41(Database issue):D344–7. Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, Tosatto SCE, Paladin L, Raj S, Richardson LJ, Finn RD, Bateman A. Pfam: The protein families database in 2021. Nucleic Acids Res. 2021;49(D1):D412–9. Letunic I, Khedkar S, Bork P. SMART: recent updates, new developments and status in 2020. Nucleic Acids Res. 2021;49(D1):D458–60. Soto LF, Li Z, Santoso CS, Berenson A, Ho I, Shen VX, Yuan S, Fuxman Bass JI. Compendium of human transcription factor effector domains. Mol Cell. 2022;82(3):514–26. Zhang Y, Yao X, Zhou H, Wu X, Tian J, Zeng J, Yan L, Duan C, Liu H, Li H, Chen K, Hu Z, Ye Z, Xu H. OncoSplicing: an updated database for clinically relevant alternative splicing in 33 human cancers. Nucleic Acids Res. 2022;50(D1):D1340–7. Cai Z, Qian ZY, Jiang H, Ma N, Li Z, Liu LY, Ren XX, Shang YR, Wang JJ, Li JJ, Liu DP, Zhang XP, Feng D, Ni QZ, Feng YY, Li N, Zhou XY, Wang X, Bao Y, Zhang XL, Deng YZ, Xie D. HPCL3s promotes hepatocellular carcinoma metastasis by activating β-catenin signaling. Cancer Res. 2018;78(10):2536–49. Xu Y, Liu W, Jiang X, Li J, Liu Q, Su F, Ruan S, Zhang Z, Tao F. Regulation of hPCL3 isoforms’ ubiquitination by TRIM21 in non-small cell lung cancer progression. Life Sci Alliance [Internet]. 2023;6(10). Available from: http://dx.doi.org/10.26508/lsa.202302060 Abdelfettah S, Boulay G, Dubuissez M, Spruyt N, Garcia SP, Rengarajan S, Loison I, Leroy X, Rivera MN, Leprince D. hPCL3S promotes proliferation and migration of androgen-independent prostate cancer cells. Oncotarget. 2020;11(12):1051–74. Wang S, Robertson GP, Zhu J. A novel human homologue of Drosophila polycomblike gene is up-regulated in multiple cancers. Gene. 2004;343(1):69–78. Cox CJ, Espinoza HM, McWilliams B, Chappell K, Morton L, Hjalt TA, Semina EV, Amendt BA. Differential regulation of gene expression by PITX2 isoforms. J Biol Chem. 2002;277(28):25001–10. Grassmeyer J, Mukherjee M, deRiso J, Hettinger C, Bailey M, Sinha S, Visvader JE, Zhao H, Fogarty E, Surendran K. Elf5 is a principal cell lineage specific transcription factor in the kidney that contributes to Aqp2 and Avpr2 gene expression. Dev Biol. 2017;424(1):77–89. Lu Y, Berenson A, Lane R, Guelin I, Li Z, Chen Y, Shah S, Yin M, Soto-Ugaldi LF, Fiszbein A, Fuxman Bass JI. A large-scale cancer-specific protein-DNA interaction network. Life Sci Alliance. 2024;7(10):e202402641. Zaika AI, Slade N, Erster SH, Sansome C, Joseph TW, Pearl M, Chalas E, Moll UM. DeltaNp73, a dominant-negative inhibitor of wild-type p53 and TAp73, is up-regulated in human tumors. J Exp Med. 2002;196(6):765–80. Fillippovich I, Sorokina N, Gatei M, Haupt Y, Hobson K, Moallem E, Spring K, Mould M, McGuckin MA, Lavin MF, Khanna KK. Transactivation-deficient p73alpha (p73Deltaexon2) inhibits apoptosis and competes with p53. Oncogene. 2001;20(4):514–22. Baek S, Goldstein I, Hager GL. Bivariate genomic footprinting detects changes in transcription factor activity. Cell Rep. 2017;19(8):1710–22. He HH, Meyer CA, Chen MW, Jordan VC, Brown M, Liu XS. Differential DNase I hypersensitivity reveals factor-dependent chromatin dynamics. Genome Res. 2012;22(6):1015–25. Keshamouni VG, Reddy RC, Arenberg DA, Joel B, Thannickal VJ, Kalemkerian GP, Standiford TJ. Peroxisome proliferator-activated receptor-gamma activation inhibits tumor progression in non-small-cell lung cancer. Oncogene. 2004;23(1):100–8. Chang TH, Szabo E. Induction of Differentiation and Apoptosis by Ligands of Peroxisome Proliferator-activated Receptor γ in Non-Small Cell Lung Cancer. Cancer Res. 2000;60(4):1129–38. Behan FM, Iorio F, Picco G, Gonçalves E, Beaver CM, Migliardi G, Santos R, Rao Y, Sassi F, Pinnelli M, Ansari R, Harper S, Jackson DA, McRae R, Pooley R, Wilkinson P, van der Meer D, Dow D, Buser-Doepner C, Bertotti A, Trusolino L, Stronach EA, Saez-Rodriguez J, Yusa K, Garnett MJ. Prioritization of cancer therapeutic targets using CRISPR-Cas9 screens. Nature. 2019;568(7753):511–6. Hamdan H, Liu H, Li C, Jones C, Lee M, deLemos R, Minoo P. Structure of the human Nkx2.1 gene. Biochim Biophys Acta. 1998;1396(3):336–48. Kimura S, Hara Y, Pineau T, Fernandez-Salguero P, Fox CH, Ward JM, Gonzalez FJ. The T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev. 1996;10(1):60–9. Li C, Cai J, Pan Q, Minoo P. Two functionally distinct forms of NKX2.1 protein are expressed in the pulmonary epithelium. Biochem Biophys Res Commun. 2000;270(2):462–8. Ryan MC, Cleland J, Kim R, Wong WC, Weinstein JN. SpliceSeq: a resource for analysis and visualization of RNA-Seq data on alternative splicing and its functional impacts. Bioinformatics. 2012;28(18):2385–7. Kahles A, Ong CS, Zhong Y, Rätsch G. SplAdder: identification, quantification and testing of alternative splicing events from RNA-Seq data. Bioinformatics. 2016;32(12):1840–7. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, Pimentel H, Salzberg SL, Rinn JL, Pachter L. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012;7(3):562–78. Ben Guebila M, Lopes-Ramos CM, Weighill D, Sonawane AR, Burkholz R, Shamsaei B, Platig J, Glass K, Kuijjer ML, Quackenbush J. GRAND: a database of gene regulatory network models across human conditions. Nucleic Acids Res. 2022;50(D1):D610–21. Additional Declarations There is NO conflict of interest to disclose. Supplementary Files SupplementaryfileS1.xlsx Supplementary Table S1: List of PTSEs for each of the 15 cancer types SupplementaryfileS2.xlsx Supplementary Table S2: List of PTSEs occurring in multiple cancer types SupplementaryfileS3.xlsx Supplementary Table S3: List of PTSEs occurring uniquely in a cancer type SupplementaryfileS4.xlsx Supplementary Table S4: List of DBD perturbing PTSEs SupplementaryfileS5.xlsx Supplementary Table S5: List of ED perturbing PTSEs SupplementaryfileS6.xlsx Supplementary Table S6: List of PTSEs correlating with TF footprinting SupplementaryfileS7.xlsx Supplementary Table S7: List of PTSEs whose corresponding transcription factors are essential for cancer cell line survival Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: revise 04 Feb, 2026 Review # 2 received at journal 30 Jan, 2026 Reviewer # 2 agreed at journal 10 Jan, 2026 Review # 1 received at journal 15 Nov, 2025 Reviewer # 1 agreed at journal 31 Oct, 2025 Reviewers invited by journal 31 Oct, 2025 Submission checks completed at journal 08 Oct, 2025 Editor assigned by journal 07 Oct, 2025 First submitted to journal 07 Oct, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7799323","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":538043429,"identity":"a4f0beb4-e251-41c2-a94e-05ccf4e3217f","order_by":0,"name":"Khalique Newaz","email":"data:image/png;base64,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","orcid":"","institution":"University of Hamburg","correspondingAuthor":true,"prefix":"","firstName":"Khalique","middleName":"","lastName":"Newaz","suffix":""},{"id":538043430,"identity":"a923e6ac-662e-4b55-bd49-80dcba1dbff7","order_by":1,"name":"Olga Tsoy","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Olga","middleName":"","lastName":"Tsoy","suffix":""},{"id":538043431,"identity":"ed2ada48-32c8-4c47-bd6f-7503c6327fc7","order_by":2,"name":"Jan Baumbach","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Jan","middleName":"","lastName":"Baumbach","suffix":""}],"badges":[],"createdAt":"2025-10-07 12:05:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7799323/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7799323/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":95668133,"identity":"1fe604d0-5613-4c7b-b724-731522e4184e","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5640524,"visible":true,"origin":"","legend":"","description":"","filename":"mainpaper.docx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/c7b291fa17b888ed6b12e9af.docx"},{"id":95668121,"identity":"eed5cfd2-0d7d-462a-848d-f40b011f53dd","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4794,"visible":true,"origin":"","legend":"","description":"","filename":"ONC202503061.json","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/1eae9a931a961496ea5c8a4f.json"},{"id":95798616,"identity":"757aebc2-c05f-498e-be52-3ec9abdcd3c3","added_by":"auto","created_at":"2025-11-13 08:17:16","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":588821,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/071412c8c6b950ac43e0ff51.xlsx"},{"id":95668128,"identity":"dd6482f1-8c42-4462-acde-98043d5dd27a","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":36435,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/017785a8384371f56adca030.xlsx"},{"id":95798363,"identity":"859cb65c-c94f-4cdb-8cc2-76f39b2b2d5a","added_by":"auto","created_at":"2025-11-13 08:16:32","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":40747,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/8213eb61af106ccbb2310f3b.xlsx"},{"id":95668132,"identity":"6cabd083-795e-4c84-8a77-f5eb741b1372","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":34820,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/f43a47e8bce346a3d362eb57.xlsx"},{"id":95668134,"identity":"1b10ad5d-5322-4f52-a51c-b75281b01656","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":32392,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/77a0bbcb88da65ce47961c4e.xlsx"},{"id":95798017,"identity":"c8b69076-9f5b-46c1-add8-8c443de77390","added_by":"auto","created_at":"2025-11-13 08:14:28","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":368412,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/35ef404b39a2cad5a509f9ae.xlsx"},{"id":95668140,"identity":"6f1eb06f-cb1e-4d97-9b97-c943bb594ebc","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":149079,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryfileS7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/a5473cf0057a128cf103b478.xlsx"},{"id":95798679,"identity":"57d4f7fe-714d-4c3a-97c8-8db3964d6c93","added_by":"auto","created_at":"2025-11-13 08:17:33","extension":"xml","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":148753,"visible":true,"origin":"","legend":"","description":"","filename":"ONC2025030610enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/d3397048500aaf91203acc28.xml"},{"id":95798160,"identity":"0fd8285f-a2fb-4692-a21b-2ff58cfa7ddc","added_by":"auto","created_at":"2025-11-13 08:15:46","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":146432,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/34ddf08d3099e03a46e3e002.jpeg"},{"id":95797322,"identity":"aaa233fa-9398-4591-b99b-f2a26a2e600d","added_by":"auto","created_at":"2025-11-13 08:03:24","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":367786,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/fd6dc5a6c9bd49798d3546f3.jpeg"},{"id":95668139,"identity":"3e57f2a6-4bff-4b9f-be59-2af959d590d1","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"jpeg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2266774,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage11.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/69968e21a7548b29d79c4b9e.jpeg"},{"id":95668145,"identity":"a6d3a1cf-807f-4695-b6aa-23d07b1ae1cf","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"jpeg","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1004365,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/a4d3b9ae21ad73de13574bc7.jpeg"},{"id":95798680,"identity":"a3340cce-63ff-46e1-acbe-0c96d420d230","added_by":"auto","created_at":"2025-11-13 08:17:33","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":77315,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/0c3f3ae53cf71e7df04c9806.png"},{"id":95668149,"identity":"3c564071-4458-45f6-8312-0a69532d7d32","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":111616,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/9a688f4f266952e83d3c4d48.png"},{"id":95668161,"identity":"d332313a-df88-4749-bc82-d0e70e0bc26f","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"jpeg","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":677085,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/0a9070787b97a2bc928a585f.jpeg"},{"id":95799099,"identity":"638513a1-7e7b-47eb-aecc-714c83c16bdf","added_by":"auto","created_at":"2025-11-13 08:18:47","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":126076,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/3c897bde2f6774099266f13b.png"},{"id":95668167,"identity":"3564083a-2f32-4d16-9f10-af1500a59aac","added_by":"auto","created_at":"2025-11-11 17:01:51","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":316216,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/a4a003c8a9daabfd189b1cd3.png"},{"id":95668157,"identity":"b46676d9-213c-49ad-9b6d-9401a7807b7e","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"jpeg","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":330277,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/01f4086e3b6fcb04541e1d93.jpeg"},{"id":95799177,"identity":"3a67451e-e97c-4997-9a13-a2b8e28421b7","added_by":"auto","created_at":"2025-11-13 08:18:59","extension":"jpeg","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":444765,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/8ff4309ad873e4c18f6119e8.jpeg"},{"id":95668160,"identity":"57fc7870-b23c-494f-9575-5d10de9010c0","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":61541,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/939a36ea79aec8a4edcf63fe.png"},{"id":95668159,"identity":"8d18e322-f957-4ce5-b486-08fc9fd3fd6c","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":82725,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/53557a7a02b5001b648380e3.png"},{"id":95668146,"identity":"7f838b3b-513d-4e0a-a2bd-c572f2bc7cd9","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":371594,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage11.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/dad060571ea05a476045f25e.png"},{"id":95668141,"identity":"bafa9e16-3238-4ddf-adee-dfe316ce6c4a","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":222110,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/d56cdca7d2780a5e4508f356.png"},{"id":95668165,"identity":"c58004d4-ca6e-4d9d-bca1-2396027c00a0","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37163,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/7b50a4e6faa9dd73136f9693.png"},{"id":95668164,"identity":"240754de-3b97-4c5f-b22e-c62aeb67522b","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42174,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/3c61242b658f8fd3236f8178.png"},{"id":95668156,"identity":"6aadab02-2a76-40da-ac47-ec380ae58494","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":84024,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/bc73186588794c46e1963be8.png"},{"id":95668144,"identity":"095920c5-734a-4938-86c0-bfe09a2f8f80","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":28,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42496,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/f9ccd293a1ae42d757353728.png"},{"id":95668163,"identity":"f83920de-a614-472c-975e-b8d09e08be68","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":29,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":54755,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/c95b43b3512a7e9929af6bae.png"},{"id":95798089,"identity":"c466cda9-ae1c-4579-b9fb-82dc4d9dd3f4","added_by":"auto","created_at":"2025-11-13 08:15:17","extension":"png","order_by":30,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":67751,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/8c4b5ac07834473b7aed425c.png"},{"id":95668155,"identity":"673b471a-201d-4e23-b9d9-17aa24da4a60","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"png","order_by":31,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":90413,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/ec571d2432fd7b4d06a62536.png"},{"id":95797830,"identity":"96379c48-fc99-4c09-8c8a-94b32c390812","added_by":"auto","created_at":"2025-11-13 08:12:02","extension":"xml","order_by":32,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":147695,"visible":true,"origin":"","legend":"","description":"","filename":"ONC2025030610structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/f8d57d1909b79a7bb75ef25e.xml"},{"id":95797324,"identity":"01aeae1e-3df0-4cd7-99c1-f629a98e3a90","added_by":"auto","created_at":"2025-11-13 08:03:25","extension":"html","order_by":33,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155684,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/4699a2d0473b0606f0d0c984.html"},{"id":95668120,"identity":"a65ccf42-fb0a-4611-a855-e71c5d1dda38","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":42037,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic illustration of AS affected exons (marked as striped boxes). 1st row: for the AS event using the second exon as AP, the first exon was marked as affected. 2nd row: for the AS event using the second last exon as AT, the last exon was marked as affected. 3rd row: for the AS event using an alternative 3’ end of an exon (AD), the part of the exon beyond alternative 3’ was marked as affected. 4th row: for the AS event using an alternative 5’ end of an exon (AA), the part of the exon before alternative 5’ was marked as affected. 5th row: for the AS event skipping an exon (ES), the corresponding exon was marked as affected. 6th row: for the AS event involving two mutually exclusive exons (MX), both exons were marked as affected.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/826ffc450791da1b4172611f.png"},{"id":95668123,"identity":"03bfb4db-aac8-4ce7-8774-309691b774e8","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":245065,"visible":true,"origin":"","legend":"\u003cp\u003eCancer type related PTSEs. \u003cstrong\u003e(A)\u003c/strong\u003e Number of cancer type related PTSEs. \u003cstrong\u003e(B)\u003c/strong\u003eDistribution of mean ΔPSI values of PTSEs. For a cancer type, the y-coordinate of a point (PTSE) represents the ΔPSI averaged over all patients with paired samples. The top of a violin bar represents the average percentage of patients in which a PTSE is consistently higher/lower in cancer samples than in paired normal samples. For example, for BLCA, this percentage is 83%. \u003cstrong\u003e(C)\u003c/strong\u003e t-SNE visualization of the cancer samples of the entire TCGA cohort. \u003cstrong\u003e(D)\u003c/strong\u003ePercentage distribution of the types of AS events.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/b9377b69321da495b5956df0.png"},{"id":95798430,"identity":"977a75c1-d9e2-46b0-bba9-22b393592108","added_by":"auto","created_at":"2025-11-13 08:16:47","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":191976,"visible":true,"origin":"","legend":"\u003cp\u003eCommon and unique PTSEs. \u003cstrong\u003e(A)\u003c/strong\u003e Number of PTSEs that occur in one or more cancer types. \u003cstrong\u003e(B)\u003c/strong\u003e Mean ΔPSI of the 39 PTSEs (y-axis) present in at least 10 cancer types (x-axis). \u003cstrong\u003e(C)\u003c/strong\u003e Number of unique PTSEs of individual cancer types, with the number of such PTSEs that are uniquely associated with patient survival denoted on top of the bars. Information about survival association was derived from Zhang et al.\u003ca href=\"https://paperpile.com/c/mLtN3s/0ZTh\"\u003e[26]\u003c/a\u003e.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/3c792555c1361be90e186965.png"},{"id":95668125,"identity":"a39aa2ee-6ab1-4086-9c1d-4b36fc491593","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":153793,"visible":true,"origin":"","legend":"\u003cp\u003ePTSEs that perturb DBDs. (A) Number of cancer type related PTSEs that perturb DBDs. (B) Distribution of the percentage of DBD residues perturbed by PTSEs. Each point represents a PTSE (with the corresponding mean ΔPSI on the y-axis) that results in a DBD perturbation (with the percentage of the perturbed DBD region on the x-axis). (C) Percentage distribution of the DBD types within all considered TFs (background) and those within PTSEs. For visualization purposes, 29 DBD types with counts of less than 5 in the background set of 1 347 TFs were grouped into one as “Others (29)”. (D) Perturbed DBD types within PTSEs. The x-axis shows the names of DBD types and the corresponding number of PTSEs separated by colon. For example, for “bHLH:8” on the second column from left denotes that bHLH is perturbed in 8 PTSEs. The y-axis shows the cancer types and the corresponding number of PTSEs separated by colon. For example, for “BLCA:8” on the first row from bottom, denotes that BLCA has 8 PTSEs.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/f7c7fb2ea565ca224a4d73d0.png"},{"id":95668130,"identity":"80121fab-f921-4a3b-823b-65848775d4f9","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":101959,"visible":true,"origin":"","legend":"\u003cp\u003ePTSEs that perturb EDs. \u003cstrong\u003e(A)\u003c/strong\u003e Number of cancer type related PTSEs that perturb EDs. \u003cstrong\u003e(B)\u003c/strong\u003e Distribution of the percentage of ED residues perturbed by PTSEs. Each point represents a PTSE (with the corresponding delta PSI on the y-axis) that results in an ED perturbation (with the percentage of the perturbed ED on the x-axis). \u003cstrong\u003e(C)\u003c/strong\u003e Number of TFs with PTSE-induced perturbations in their DBD regions, ED regions, both DBD and ED regions, or other protein regions.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/8013ee103debc6f69248bb1e.png"},{"id":95798044,"identity":"e2c3c222-122f-4b64-b3df-50a10132ecc6","added_by":"auto","created_at":"2025-11-13 08:14:51","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":105343,"visible":true,"origin":"","legend":"\u003cp\u003ePTSEs correlated with footprinting. Each point represents one of the 1 948 PTSE-cancer type combinations (Table 2). Note that one PTSE-cancer type combination can have multiple footprinting depth/flanking accessibility values because of the presence of multiple DNA motifs for the corresponding TF (Section 2.6). In this case, the maximum correlation value (for each of footprinting depth and flanking accessibility) was taken to generate this figure.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/00debe1f613a7feb0298db72.png"},{"id":95668143,"identity":"2cf16315-748c-43dd-b894-f7159d1a0f19","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":480772,"visible":true,"origin":"","legend":"\u003cp\u003eCancer dependency of TFs involved in PTSEs. \u003cstrong\u003e(A)\u003c/strong\u003e Distribution of Chronos scores of all 2 505 TF-cancer type combinations (Section 2.7). Each point represents a TF-cancer type combination with x-axis representing the maximum ΔPSI value among all PTSEs affecting the TF and the y-axis representing the minimum Chronos score of the TF among all cell lines of the cancer type. \u003cstrong\u003e(B)-(M)\u003c/strong\u003e Chronos scores of individual cancer types, with x-axis showing the cell line and the y-axis showing the top 1% of PTSEs with maximum absolute mean ΔPSI value (denoted within brackets).\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/7d2709d60f587a5ff281118c.png"},{"id":95804593,"identity":"0d1c56d0-21b9-4f5d-bc41-1bccab281aec","added_by":"auto","created_at":"2025-11-13 08:38:40","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2033304,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/d953eacb-1cf1-41b6-bec4-038c5f98f400.pdf"},{"id":95798660,"identity":"048ad134-cf43-486e-96b6-c7d15c3c0ee9","added_by":"auto","created_at":"2025-11-13 08:17:28","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":588821,"visible":true,"origin":"","legend":"Supplementary Table S1: List of PTSEs for each of the 15 cancer types","description":"","filename":"SupplementaryfileS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/06afb239c851fae460ad8b94.xlsx"},{"id":95798629,"identity":"3722a16c-9efa-4c24-ab39-7d3eb1196b01","added_by":"auto","created_at":"2025-11-13 08:17:21","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":36435,"visible":true,"origin":"","legend":"Supplementary Table S2: List of PTSEs occurring in multiple cancer types","description":"","filename":"SupplementaryfileS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/a9889a38a6816687073309f5.xlsx"},{"id":95798488,"identity":"4fba3297-0003-4380-83eb-870fa07ba127","added_by":"auto","created_at":"2025-11-13 08:16:54","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":40747,"visible":true,"origin":"","legend":"Supplementary Table S3: List of PTSEs occurring uniquely in a cancer type","description":"","filename":"SupplementaryfileS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/e291229a9c5ae0f9574a82c2.xlsx"},{"id":95797821,"identity":"515104bf-4b88-4dc1-8d39-597388c35db2","added_by":"auto","created_at":"2025-11-13 08:11:42","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":34820,"visible":true,"origin":"","legend":"Supplementary Table S4: List of DBD perturbing PTSEs","description":"","filename":"SupplementaryfileS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/a5dc0252d9154c08e741535c.xlsx"},{"id":95668136,"identity":"e32d1a86-5d79-4394-9f5c-db90285b23a5","added_by":"auto","created_at":"2025-11-11 17:01:49","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":32392,"visible":true,"origin":"","legend":"Supplementary Table S5: List of ED perturbing PTSEs","description":"","filename":"SupplementaryfileS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/366b59565f1fade77eda70c2.xlsx"},{"id":95798105,"identity":"85337388-b87f-42af-9f43-77ce083ec97e","added_by":"auto","created_at":"2025-11-13 08:15:29","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":368412,"visible":true,"origin":"","legend":"\u003cp\u003eSupplementary Table S6: List of PTSEs correlating with TF footprinting\u003c/p\u003e","description":"","filename":"SupplementaryfileS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/35713f93ffce7adc5859497f.xlsx"},{"id":95668166,"identity":"f58fe7ce-0a89-41b8-8bde-0413e14fa756","added_by":"auto","created_at":"2025-11-11 17:01:50","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":149079,"visible":true,"origin":"","legend":"Supplementary Table S7: List of PTSEs whose corresponding transcription factors are essential for cancer cell line survival","description":"","filename":"SupplementaryfileS7.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7799323/v1/cc326ef05e1ba411eacc2450.xlsx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose.","formattedTitle":"Aberrant splicing in human cancer has a large-scale functional impact on transcription factors","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eAlternative splicing (AS) results in multiple different messenger ribonucleic acids (mRNAs) originating from a single gene[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. AS is associated with normal cellular processes, including tissue development and identity[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Normal AS patterns are perturbed in diseases[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], including in cancer[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Large-scale RNA sequencing (RNAseq)-based evidence for aberrant AS in cancer has been collected. For example, the cancer genome atlas (TCGA) SpliceSeq database contains RNAseq-based quantification of AS events across 33 cancer types[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. TCGA SpliceSeq has been used to explore cancer-specific AS[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. For example, Yongsheg et al.[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e] investigated relationships between somatic mutations in cancer and cancer-specific AS patterns, and found that different mutations in the same gene showed distinct AS perturbations. Junyi et al.[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] exploited the hypothesis that RNA-binding proteins (RBPs) regulate cancer-specific AS patterns and proposed a computational pipeline to identify RBPs that are cancer-specific AS regulators. Similarly, Cheng et al.[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] investigated whether abnormal AS of splicing factors could explain cancer-specific AS patterns.\u003c/p\u003e\u003cp\u003eAberrant AS in cancer has also been shown to impact an important group of proteins called transcription factors (TFs)[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. TFs bind DNA to regulate transcription of genes[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Cancer-specific dysregulation of TFs perturbs normal transcriptional programs, which could lead to a sustained tumor state[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Hence, TFs have been considered potential cancer therapeutic targets[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. TF dysregulations could be due to cancer-specific signaling cascades that perturb TF expression or changes in the corresponding gene structure via chromosomal translocations/point mutations[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Some evidence of cancer-specific AS effects on TFs has also been documented[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. For example, AS-induced alternative protein isoforms of T Cell Factor 4 (\u003cem\u003eTCF4\u003c/em\u003e) have been implicated in proliferations of non-small cell lung cancer cells[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] and hepatocellular carcinoma[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Other examples are reviewed elsewhere[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Despite existing evidence, a comprehensive systematic analysis of how cancer-specific AS affects TFs has been missing. Such an analysis could reveal how TF-affecting AS events drive cancer progression, which could then be used to identify therapeutic targets and strategize drug design[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eHere, TCGA SpliceSeq was used to systematically explore the effect of cancer-specific AS events across all known 1 639 human TFs. 15 cancer types that had AS data for enough paired (cancer and healthy) patient samples were used. For each cancer-type, AS events showing statistically significant differences between paired cancer and normal samples were considered as perturbed. 2 118 TF-affecting perturbed AS events covering 718 TFs across 14 cancer types were found. Comparisons across cancer types revealed several unique cancer type-specific, as well as common, TF-affecting perturbed AS events. In 189 of the 718 (~\u0026thinsp;26%) TFs, at least one perturbed AS event affects TF functional domains. This provides direct evidence that cancer-specific AS events have a functional impact on TFs. Additionally, we found indirect evidence of the functional impact of cancer-specific AS events on TFs in two ways. First, we found that the expression patterns of \u0026gt;\u0026thinsp;100 AS events covering\u0026thinsp;\u0026gt;\u0026thinsp;70 of the 718 TFs have significant correlations with the existing sample-matched TF footprinting data[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], implying a possible dependence of the DNA binding and gene regulatory activities of the affected TFs on the corresponding perturbed AS events. Second, utilizing existing cancer dependency data[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e], we found that many (39) TFs affected by perturbed AS events are also essential for cancer progression in cancer cell lines. Overall, our findings highlight a large-scale functional impact of cancer-specific AS on TFs. The collected evidence could guide future experiments to understand cancer molecular mechanisms or identify therapeutic targets.\u003c/p\u003e"},{"header":"2. Data and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Alternative splicing data\u003c/h2\u003e\u003cp\u003eFrom the TCGA database[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e] (accessed Feb. 2025), 15 cancer types were selected, where each cancer type had at least 10 patients with primary tumor (cancer) and patient-matched control (healthy) samples (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). For these paired samples, the corresponding AS data were downloaded from TCGA SpliceSeq[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. This data contains information about seven types of AS events: Alternative Acceptor (AA), Alternative Donor (AD), Alternative Promoter (AP), Alternative Terminator (AT), Exon Skip (ES), Mutually Exclusive Exons (ME), and Retained Intron (RI). To quantify AS events in a TCGA sample, TCGA SpliceSeq provides Percent Spliced In (PSI) values for AS events covered by at least eight RNA-seq reads; for the rest of the AS events, a \u0026ldquo;null\u0026rdquo; value is provided. For each cancer type, we only considered events with PSI values for at least 10 paired samples.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003ePatient counts across 15 cancer types in TCGA SpliceSeq.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCancer type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTCGA cancer code\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e# of patients with paired samples\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e# of cancer samples\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e# of AS events with\u003c/p\u003e\u003cp\u003evalues for at least\u003c/p\u003e\u003cp\u003e10 paired samples\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBladder urothelial carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBLCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e406\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e42 343\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBreast invasive carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eBRCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e96\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1094\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e58 913\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eColon adenocarcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eCOAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e39\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e458\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e48 843\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEsophageal carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eESCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e180\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e44 778\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHead and neck squamous cell carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eHNSC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e501\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e51 442\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKidney chromophobe\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eKICH\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e66\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e50 059\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKidney renal clear cell carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eKIRC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e71\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e533\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e57 286\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKidney renal papillary cell carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eKIRP\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e32\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e290\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e52 337\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLiver hepatocellular carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLIHC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e371\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e43 753\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLung adenocarcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLUAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e514\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e51 436\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLung squamous cell carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLUSC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e48\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e501\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e57 040\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eProstate adenocarcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ePRAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e497\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e55 799\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStomach adenocarcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSTAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e33\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e415\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e54 446\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eThyroid carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTHCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e505\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e56 824\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUterine corpus endometrial carcinoma\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eUCEC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e545\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e36 585\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eFor each cancer type, significantly perturbed AS events were detected as follows. For an AS event, only the patients with PSI values for both normal and cancer samples were detected. Considering these patients, a two-sided paired Wilcoxon Signed Rank test was performed to compare PSI values from normal samples and the paired cancer samples using the \u0026ldquo;\u003cem\u003ewilcox.test\u003c/em\u003e\u0026rdquo; function in \u003cem\u003eR\u003c/em\u003e to obtain a p-value. The p-values across all AS events were corrected using the Benjamini Hochberg procedure to obtain the corresponding q-values. An AS event was considered as significantly perturbed if and only if (1) its q-value was \u0026lt;\u0026thinsp;0.05 and (2) the difference between the means of normal and cancer PSI values was \u0026gt;\u0026thinsp;5%.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Transcription factors\u003c/h2\u003e\u003cdiv id=\"Sec5\" class=\"Section3\"\u003e\u003ch2\u003e2.2.1 Curation and mapping of genomic coordinates\u003c/h2\u003e\u003cp\u003eAll 1 639 curated human TFs were obtained from Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e of the Lambert et al.[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] study. For these TFs, canonical protein sequences from UniProt were collected. Henceforth, by the term \u0026ldquo;UniProt protein sequence\u0026rdquo; we mean the \u0026ldquo;canonical UniProt protein sequence\u0026rdquo;. UniProt IDs and Ensembl transcripts were matched using UniProt. Because one TF can have multiple Ensembl transcripts, one transcript per TF was selected as follows. Given a TF, its UniProt sequence was pairwise globally aligned with each of the corresponding Ensembl transcripts. Protein sequences of the transcripts were constructed using the release 113 of Genome Reference Consortium Human Build 37 (GRCh37) Ensembl GTF file \u0026ldquo;Homo_sapiens.GRCh37.dna.primary_assembly.fa.gz\u0026rdquo;. Then, the transcript that matched the TF protein sequence with 100% identity and no gaps was selected. For the 1 469 TFs with matched transcripts as per the above criterion, genomic coordinates were mapped onto their protein sequences using GRCh37. Note that GRCh37 was used instead of GRCh38 because TCGA SpliceSeq used GRCh37 to map genomic coordinates of the AS events. Genomic coordinates of AS events provided by TCGA SpliceSeq were then used to mark parts of the TF protein sequences affected by AS events.\u003c/p\u003e\u003cp\u003eSpecifically, for each of the six types of AS events, i.e., AA, AD, AP, AT, ES, and ME, the following was done. Note that RI AS events can not be mapped to TFs as they do not overlap with annotated protein-coding regions defined by GRCh37. For AA, AD, ES, or ME, all sequence positions of a TF whose start and stop genomic coordinates lie within the start and stop coordinates of the affected exons of an AS event were marked as affected. For AP, all sequence positions of a TF whose start genomic coordinates lie before the start position of the affected exon were marked as affected. For AT, all sequence positions of a TF whose stop genomic coordinates lie beyond the stop position of the affected exon were marked as affected.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section3\"\u003e\u003ch2\u003e2.2.2 Mapping of DNA-binding and effector domains\u003c/h2\u003e\u003cp\u003eUniProt collects DNA-binding domain (DBD) information of proteins using PROSITE[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], Pfam[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e], or SMART[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. This DBD information was used to annotate which sequence positions of TFs belong to a DBD region. Specifically, given the information about a TF in UniProt, if present, the following feature table entries were considered as a DBD: (1) entries \u0026ldquo;DNA_BIND\u0026rdquo; or \u0026ldquo;ZN_FING\u0026rdquo; or (2) entry \u0026ldquo;DOMAIN\u0026rdquo; with domain names 'bHLH', 'bZIP', or 'MADS-box'. Of the 1 469 TFs with a matching Ensembl transcript (Section \u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003e2.2.1\u003c/span\u003e), 1 347 had at least one sequence position belonging to a DBD. TF effector domain (ED) information was taken from Supplementary Tables S2 and S3 of the Soto et al.[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] study. This study looked for ED evidence across 1 639 human TFs in the literature and manually curated the largest collections of 924 EDs across 594 TFs. For these TFs information about the UniProt sequence (the corresponding isoform) positions of EDs are provided. This ED information was used to annotate which UniProt sequence positions of TFs belong to an ED region. Of the 1 469 TFs (Section \u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003e2.2.1\u003c/span\u003e), 460 had at least one sequence position belonging to an ED region, covering 730 EDs.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.3 tSNE application\u003c/h2\u003e\u003cp\u003eAcross all 15 cancer types, 2 278 TF AS events were identified that were present in each cancer type. Among these, 1 011 TF AS events were present as perturbed TF splicing events (PTSEs) in at least one cancer type. PSI patterns of these 1 011 AS events for all 6 876 cancer samples across the 15 cancer types were used as input to the \u0026ldquo;\u003cem\u003eRtsne\u003c/em\u003e\u0026rdquo; package of R to obtain dimensionally reduced representations of the cancer samples.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Hypergeometric test of DBD perturbations\u003c/h2\u003e\u003cp\u003eAmong the 1 347 TFs with DBD annotation (Section \u003cspan refid=\"Sec6\" class=\"InternalRef\"\u003e2.2.2\u003c/span\u003e), 1 464 occurrences of DBDs were found, which were taken as the background size. Among all PTSEs across all cancer types, 146 unique DBDs were present, which were taken as the sample size. For each of the 27 DBD-types perturbed in at least one PTSE, their counts in the sample and the background were used to compute the corresponding over-representations via the hypergeometric test to obtain 27 p-values. Similarly, for each of the 27 DBD-types, their counts in the sample and the background were used to compute the corresponding under-representations via the hypergeometric test to obtain 27 p-values. The 54 hypergeometric p-values were corrected using the Benjamini Hochberg procedure to obtain the corresponding q-values.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e2.5 Alternative splicing events significantly associated with patient survival\u003c/h2\u003e\u003cp\u003eAssociations between AS events from TCGA SpliceSeq and cancer patient survival were obtained from the OncoSplicing database (\u0026ldquo;SpliceSeq_info_Survival.csv.gz\u0026rdquo;) developed in Zhang et al.[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. For a cancer type, OncoSplicing contains AS events that correlate with clinical outcomes according to the Cox proportional hazards regression. All AS events with survival association p-values of less than 0.05 (reported in the column named \u0026lsquo;Pvalue_HR\u0026rsquo; of \u0026ldquo;SpliceSeq_info_Survival.csv.gz\u0026rdquo;) were considered as significant and selected for our study.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e2.6 Correlation of PSI values of PTSEs and TF footprinting\u003c/h2\u003e\u003cp\u003eTF footprinting data were taken from Supplementary Table \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e of Corces et al.[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e], which covers 410 TCGA cancer samples and 843 TFs across 23 cancer types. Corces et al. provides footprinting depth and flanking accessibility values for 1 781 TF-DNA motif pairs, with multiple DNA motifs for a given TF. For a cancer type, a correlation between PSI values of a PTSE and footprinting values of a TF-DNA motif pair was measured if and only if there were at least 10 cancer samples for which the PTSE had a non-null PSI value in TCGA SpliceSeq and footprinting data in Corces et al. At least 10 cancer samples were considered to retain power in the correlation analysis. 769 PTSEs across 12 cancer types fulfilled this criterion (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eOverlap between TCGA SpliceSeq and Corces et al.\u003csup\u003e19\u003c/sup\u003e.\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTCGA\u003c/p\u003e\u003cp\u003ecancer code\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003e# of cancer\u003c/p\u003e\u003cp\u003esamples\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003e# of PTSEs with PSI values for at least 10 cancer samples\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003e# of TFs within PTSEs with PSI values for at least 10 cancer samples\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBLCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e42\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e28\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBRCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e296\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e132\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCOAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e38\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e176\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e88\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKIRC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e245\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e116\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKIRP\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e33\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e150\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e84\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLIHC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e37\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLUAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e20\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e198\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e93\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLUSC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e299\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e135\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePRAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e26\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e124\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e67\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSTAD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e96\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e56\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTHCA\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e14\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e144\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e78\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUCEC\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e13\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e85\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e52\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eGiven a cancer type, for each PTSE selected as per the above criterion, we measured the correlation between its PSI values and footprinting depth of the corresponding TF-DNA motif pairs. Over all 12 cancer types this resulted in 3 676 correlation pairs. Similarly, we obtained 3 676 correlation pairs corresponding to flanking accessibility. We calculated the empirical p-value based on the data randomization. For each of the 3 676*2\u0026thinsp;=\u0026thinsp;7 352 pairs, 100 random correlations were computed by randomly shuffling the PSI values and the footprinting data. The fraction of random correlations higher than the real correlation was taken as the empirical p-value. The resulting 7 352 p-values were corrected using the Benjamini Hochberg procedure to obtain the corresponding q-values. A q-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered significant.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e2.7 Dependency profiles of TFs associated with PTSEs\u003c/h2\u003e\u003cp\u003eDependency profiles of cancer cell lines were downloaded from the depmap[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e] portal (\u0026ldquo;\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://depmap.org/portal/data_page/?tab=customDownloads\u003c/span\u003e\u003cspan address=\"https://depmap.org/portal/data_page/?tab=customDownloads\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e\u0026rdquo;)\u003c/span\u003e, which included all available cell lines and genes. This data cover 12 cancer types considered in our study: BLCA, BRCA, COAD, HNSC, KIRC, LIHC, LUAD, LUSC, PRAD, STAD, THCA, and UCEC. For these 12 cancer types, cell lines corresponding the following terminologies (in the order of the cancer types mentioned above) from the \u0026ldquo;lineage_3\u0026rdquo; column of the data was used: \u0026ldquo;Bladder Urothelial\u0026rdquo;, \u0026ldquo;Breast Invasive Carcinoma\u0026rdquo;, \u0026ldquo;Colorectal Adenocarcinoma\u0026rdquo;, \u0026ldquo;Head and Neck Squamous Cell Carcinoma\u0026rdquo;, \u0026ldquo;Renal Clear Cell Carcinoma\u0026rdquo;, \u0026ldquo;Hepatocellular Carcinoma\u0026rdquo;, \u0026ldquo;Lung Adenocarcinoma\u0026rdquo;, \u0026ldquo;Lung Squamous Cell Carcinoma\u0026rdquo;, \u0026ldquo;Prostate Adenocarcinoma\u0026rdquo;, \u0026ldquo;Stomach Adenocarcinoma\u0026rdquo;, \u0026ldquo;Thyroid Cancer\u0026rdquo;, \u0026ldquo;Endometrial Carcinoma\u0026rdquo;. The PTSEs across all of the above 12 cancer types cover 688 TFs. Of the 688 TFs, for 669 TFs across the 12 cancer types, there is dependency data with respect to at least one cancer cell line, totaling 2 505 TF-cancer type combinations.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results and discussion","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 Perturbed TF splicing events in individual cancer types\u003c/h2\u003e\n \u003cp\u003eGiven a cancer type, all AS events showing statistically significant (q-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and a difference of \u0026gt;\u0026thinsp;5% between the means of PSI values from normal and cancer samples) PSI differences were considered as perturbed AS events (Section \u003cspan class=\"InternalRef\"\u003e2.1\u003c/span\u003e). From these, all AS events corresponding to TFs (Section \u003cspan class=\"InternalRef\"\u003e2.2\u003c/span\u003e) were considered as perturbed TF splicing events (PTSEs).\u003c/p\u003e\n \u003cp\u003eOver all cancer types, 718 TFs were affected by at least one PTSE. Many PTSEs were found for each cancer type, except for ESCA that showed no perturbed AS event and thus no PTSE. No perturbed AS event for ESCA could be due to the low number of samples (13; see Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e) that do not provide enough statistical power to detect significant differences between cancer and normal samples with the current q-value cutoff. The number of PTSEs showed large variations across cancer types, with the highest number (750) of PTSEs for LUSC (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eA). On average over all cancer types, ~\u0026thinsp;1.8 PTSEs came from every TF, implying that no TF dominates PTSEs. PTSE numbers within cancer types do not seem to correlate with the corresponding numbers of paired normal and cancer samples (Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e). For example, LIHC, LUAD, LUSC, and PRAD, had similar numbers (50, 51, 48, and 51 respectively) of paired samples, but their PTSE numbers varied substantially: 247, 461, 750, and 310, respectively. Overall, these results indicate that not only are there significant TF AS differences between normal and cancer samples within a cancer type, but these differences also vary considerably across cancer types.\u003c/p\u003e\n \u003cp\u003eFor many PTSEs within a cancer type, their PSI values of cancer samples were higher/lower than their PSI values of the paired normal samples for a majority of patients (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eB). For example, the PSI values for the AS event \u003cem\u003eHMGA2_22880_AT\u003c/em\u003e (here, \u0026ldquo;HMGA2\u0026rdquo; is the gene symbol, \u0026ldquo;22880\u0026rdquo; is the AS event number provided by the TCGASpliceSeq database, and \u0026ldquo;AT\u0026rdquo; is the type of AS event) were higher in cancer than normal samples for 52 out of the 57 paired samples of THCA with a mean \u0026Delta;PSI of around 0.44 (Supplementary Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e). Here, \u0026Delta;PSI was defined as the difference between the PSI value of a cancer sample and the paired normal sample, with mean \u0026Delta;PSI denoting the average \u0026Delta;PSI over all paired samples of a given cancer type. The above results imply cancer type relevance of PTSEs. To further check whether this is the case, we used t-distributed Stochastic Neighbor Embedding (tSNE) to visualize cancer samples of the entire cohort of the considered cancer types, based on PSI patterns of all 1 011 AS events that were present in each of the 15 cancer types and were present as PTSEs in at least one cancer type (Section \u003cspan class=\"InternalRef\"\u003e2.3\u003c/span\u003e). We found that cancer type-related samples were generally clustered together (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eC), implying cancer tissue specificity of the PSI patterns of PTSEs. Similar results have been shown using PSI patterns of all AS events in cancer[\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e]. Interestingly, although we found large variations in PTSEs with respect to their numbers and PSI patterns across cancer types, the proportions of different types of AS events within PTSEs were similar across cancer types. AP, AT, and ES dominate PTSEs (Fig. \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eD), which has also been found in existing studies using all AS events[\u003cspan class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e9\u003c/span\u003e].\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Common and unique PTSEs of individual cancer types\u003c/h2\u003e\n \u003cp\u003eOf the total 2 118 PTSEs, a majority (~\u0026thinsp;61%, i.e., 1 302 PTSEs) occurred in more than one cancer type (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA and Supplementary Table \u003cspan class=\"InternalRef\"\u003eS2\u003c/span\u003e). 39 PTSEs occurred in at least 10 cancer types (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eB). An example PTSE is \u003cem\u003ePHD19_87401_AT\u003c/em\u003e that affects the PHD Finger Protein 19 (\u003cem\u003ePHD19\u003c/em\u003e) gene. This PTSE was upregulated in cancer samples in 12 cancer types, and corresponds to a known PHF19 isoform (hPCL3s) that is related to cancer progression of multiple cancer types[\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan class=\"CitationRef\"\u003e30\u003c/span\u003e]. We also found many unique PTSEs (816 in total) for individual cancer types that varied across cancer types, with BRCA having the most (151) and BLCA having the least (6) of them (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eC). Interestingly, at least one of the unique PTSE is survival-associated with the corresponding cancer type (Section \u003cspan class=\"InternalRef\"\u003e2.5\u003c/span\u003e), but not to any of the other cancer types (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eC and Supplementary Table \u003cspan class=\"InternalRef\"\u003eS3\u003c/span\u003e). For example, PTSE \u003cem\u003ePITX2_70356_AP\u003c/em\u003e affects the Pituitary homeobox 2 (\u003cem\u003ePITX2\u003c/em\u003e) gene that results in the PITX2 isoform named PITX2D[\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e]. \u003cem\u003ePITX2_70356_AP\u003c/em\u003e was upregulated in cancer samples compared to normal samples in BLCA with a mean \u0026Delta;PSI of 0.27, and was found to be associated with disease-free interval of BLCA with a hazard ratio of 0.31 (Supplementary Table \u003cspan class=\"InternalRef\"\u003eS3\u003c/span\u003e). This means that an increase in the inclusion of this PTSE indicates a decreased risk of BLCA recurrence. PITX2 isoforms have been shown to be crucial for developmental processes, with PITX2D inhibiting the transcriptional activities of two other PITX2 isoforms by physically interacting with them[\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e]. This might explain why the increased expression of \u003cem\u003ePITX2_70356_AP\u003c/em\u003e, which potentially leads to an increased expression of PITX2D, correlates with a decrease in the risk of BLCA recurrence. These results suggest that unique PTSEs could be relevant to understand the molecular mechanisms of disease progression of individual cancer types.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3 PTSEs perturb DNA-binding domains\u003c/h2\u003e\n \u003cp\u003eDBDs are important functional components of TFs, which enable TFs to recognize and bind the specific DNA sequences related to their target genes[\u003cspan class=\"CitationRef\"\u003e11\u003c/span\u003e]. AS events affecting DBDs can impact the DNA-binding ability of the resulting TF isoforms. To capture this, we evaluated whether PTSEs affect DBD regions of TFs (Section \u003cspan class=\"InternalRef\"\u003e2.2\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eOver all cancer types, DBD regions of 135 TFs (i.e., ~\u0026thinsp;10% of all 1 347 TFs with a DBD information; Section \u003cspan class=\"InternalRef\"\u003e2.2\u003c/span\u003e) were perturbed by at least one PTSE (Supplementary Table \u003cspan class=\"InternalRef\"\u003eS4\u003c/span\u003e). Similar to the numbers of all PTSEs (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA), numbers of DBD perturbing PTSEs show large variations across cancer types, with the highest number (87) for LUSC and the lowest number (8) for BLCA (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eA). Many of such PTSEs affect large portions of DBD regions within protein sequences of TFs (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eB), and thus could have a strong functional impact. An interesting example is the PTSE \u003cem\u003eELF5_14951_AT\u003c/em\u003e, which perturbs the Erythroblast Transformation Specific (ETS) domain of E74-like transcription factor 5 (ELF5). This PTSE has higher PSI values in cancer samples than in normal samples of kidney cancer: KICH, KIRP, and KIRC (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eB). \u003cem\u003eELF5_14951_AT\u003c/em\u003e seldom shows large differences among normal vs. cancer samples in other cancer types (Supplementary Table \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e), implying its potential functional relevance in kidney cancer. Interestingly, ELF5 has been shown to be a lineage-specific TF of principal cells that are key cells involved in kidney development[\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e]. Overall, these results indicate that PTSEs could have functional relevance in cancer phenotype development.\u003c/p\u003e\n \u003cp\u003eWe next evaluated whether certain DBD types are significantly over/under-represented within PTSEs. Specifically, we quantified the number of DBDs within PTSEs of any cancer type compared to the number of DBDs over all considered TFs, using the hypergeometric test (Section \u003cspan class=\"InternalRef\"\u003e2.4\u003c/span\u003e). The proportions of DBDs among all considered TFs and those present in PTSEs were similar, with one significant exception (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eC). That is, the Homeobox domain was significantly under-represented (q-value\u0026thinsp;\u0026lt;\u0026thinsp;10\u003csup\u003e\u0026minus;\u0026thinsp;06\u003c/sup\u003e) in PTSEs. Under-representation of the Homeobox domain in cancer was also observed in a recent study that detected 1 350 TF-DNA interactions among 265 TFs and cloned promoters of 108 cancer-related genes[\u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e]. These results mean that the Homeobox domain is not only less relevant for cancer-related gene regulations in general, but it is also less affected by cancer-related AS events. Within individual cancer types, the DBD types show different perturbation patterns (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eD). Some DBD types (i.e., bHLH, bZIP, C2H2-type, IRF tryptophan pentad repeat, PHD-type, and THAP-type) were perturbed in multiple (\u0026gt;\u0026thinsp;12) cancer types, while others (i.e., A.T hook, CCHHC-type, and Homeobox) were perturbed in selected (\u0026lt;\u0026thinsp;3) cancer types. These results could reflect the underlying AS-induced shared and unique TF-DNA binding perturbations among cancer types.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003e3.4 PTSEs perturb effector domains\u003c/h2\u003e\n \u003cp\u003eBesides DBDs, effector domains (EDs) also constitute important functional components of TFs that enable activation or repression of the target genes[\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]. Thus, AS events affecting EDs can also impact the functionality of the resulting TF isoforms. To capture this, we evaluated whether PTSEs affect ED regions of TFs (Section \u003cspan class=\"InternalRef\"\u003e2.2\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eSimilar to the numbers of DBD perturbing PTSEs (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eA), numbers of ED perturbing PTSEs show large variations across cancer types, with the highest number (51) for KIRC and the lowest number (14) for BLCA (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eA and Supplementary Table \u003cspan class=\"InternalRef\"\u003eS5\u003c/span\u003e). Similar to DBD perturbing PTSEs, many ED perturbing PTSEs affect large portions of ED regions within TF protein sequences (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eB). For example, the PTSE \u003cem\u003eTP73_321_AP\u003c/em\u003e perturbs the activator domain (AD) of TP73 in LUSC. This PTSE has higher PSI values in cancer samples than in normal samples of LUSC (Fig. \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003eB). \u003cem\u003eTP73_321_AP\u003c/em\u003e leads to the \u0026Delta;Np73 isoform of TP73, which is known to be overexpressed in tumors compared to normal conditions[\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e] and has been shown to fail in inducing apoptosis and cell cycle arrest[\u003cspan class=\"CitationRef\"\u003e35\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eOver all cancer types, 149 PTSEs perturb ED regions across 84 TFs (i.e., ~\u0026thinsp;18% of all 460 TFs with ED information; Section \u003cspan class=\"InternalRef\"\u003e2.2\u003c/span\u003e). This proportion is almost twice the proportion of TFs (~\u0026thinsp;10%) whose DBD regions were perturbed by PTSEs. This implies that PTSEs more frequently affect the transcriptional activity of TFs than their DNA binding ability. Of all 718 TFs that are perturbed by at least one PTSE, 189 (~\u0026thinsp;26%) involve a DBD or an ED, of which 29 TFs show perturbations in both ED and DBD regions. These results imply that AS causes large-scale functional perturbation of TFs in cancer.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003e3.5 Alternative splicing patterns of PTSEs correlate with TF footprinting\u003c/h2\u003e\n \u003cp\u003eTo determine potentially AS-driven TF binding/activity patterns, we correlated the PSI values of PTSEs with the TF footprinting of the corresponding TFs in cancer samples of individual cancer types (Section \u003cspan class=\"InternalRef\"\u003e2.6\u003c/span\u003e). To do this, the chromatin accessibility data from Corces et al.[\u003cspan class=\"CitationRef\"\u003e19\u003c/span\u003e] was used. Corces et al. used assay for transposase-accessible chromatin using sequencing (ATAC-seq) to determine chromatin occupancy of TFs in TCGA cancer samples, quantified via \u0026ldquo;footprinting depth\u0026rdquo; and \u0026ldquo;flanking accessibility\u0026rdquo;. Intuitively, footprinting depth measures the inverse of the extent to which the binding of a TF to a chromatin region protects the corresponding TF-binding DNA motif from DNase[\u003cspan class=\"CitationRef\"\u003e36\u003c/span\u003e]. Thus, low footprinting depth means high TF binding, and vice versa. In contrast, flanking accessibility measures the accessibility of the DNA region adjacent to TF-binding DNA motifs, which has been associated with TF activity[\u003cspan class=\"CitationRef\"\u003e37\u003c/span\u003e]. Thus, high flanking accessibility could mean high TF activity, and vice versa.\u003c/p\u003e\n \u003cp\u003eWe found 122 PTSEs covering 77 TFs across 11 cancer types (146 unique PTSE-cancer type combinations) that showed significant (q-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05) correlation with footprinting depth, flanking accessibility, or both (Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e and Supplementary Table \u003cspan class=\"InternalRef\"\u003eS6\u003c/span\u003e). Many of these PTSEs showed high mean \u0026Delta;PSI values. For example, the PTSE \u003cem\u003ePPARG_63413_AP\u003c/em\u003e affecting Peroxisome Proliferator-Activated Receptor Gamma (PPARG) in LUSC is significantly negatively correlated with footprinting depth and positively correlated with flanking accessibility. This could imply that, in LUSC, with an increase in the proportion of \u003cem\u003ePPARG_63413_AP\u003c/em\u003e there is an increase in the DNA binding as well as the gene regulatory activity of PPARG. Because \u003cem\u003ePPARG_63413_AP\u003c/em\u003e shows a mean \u0026Delta;PSI of -0.42 in LUSC, DNA binding and gene regulatory activities of PPARG is perhaps higher in the normal tissue than in LUSC. Interestingly, previous in vivo[\u003cspan class=\"CitationRef\"\u003e38\u003c/span\u003e] and in vitro[\u003cspan class=\"CitationRef\"\u003e39\u003c/span\u003e] studies have shown that an increase in the PPARG activity slows down the tumor growth rate in non-small-cell lung cancer. Overall, this analysis identified PTSEs within cancer types whose inclusion rates are correlated with the activities of the corresponding TFs.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e\n \u003ch2\u003e3.6 PTSEs include TFs showing cancer dependencies\u003c/h2\u003e\n \u003cp\u003eLoss-of-function screens in cancer cell lines, e.g., using Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)[\u003cspan class=\"CitationRef\"\u003e40\u003c/span\u003e], have identified genes essential for cancer cell line survival. Because many TFs have been associated with the sustained cancer cell state[\u003cspan class=\"CitationRef\"\u003e12\u003c/span\u003e], we evaluated the extent to which the TFs within PTSEs show cancer dependency, using the CRISPR-Cas9 screening data from The Cancer Dependency Map (DepMap, Broad (2025). DepMap Public 25Q2. Dataset. depmap.org)[\u003cspan class=\"CitationRef\"\u003e20\u003c/span\u003e] (Section \u003cspan class=\"InternalRef\"\u003e2.7\u003c/span\u003e). Given a gene and a cancer cell line, DepMap provides gene perturbation effects based on Chronos score. A Chronos score of 0 or above indicates that the gene is not essential for the cancer cell proliferation, while a Chronos score of below 0 indicates the gene is essential for the cancer cell line proliferation, with the more negative value indicating more essentiality.\u003c/p\u003e\n \u003cp\u003eWe found 669 TFs covered by PTSEs across 12 cancer types that had cancer dependency data for at least one cancer cell line (Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eA). 646 of the 669 TFs showed a negative Chronos score in cell lines associated with at least one cancer type (Supplementary Table \u003cspan class=\"InternalRef\"\u003eS7\u003c/span\u003e). 39 of these TFs showed high (Chronos score \u0026lt; -1) dependency in a cancer cell line (Fig. \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eB-M). An interesting example is the PTSE \u003cem\u003eNKX2-1_27289_ES\u003c/em\u003e affecting Homeobox protein Nkx-2.1 (NKX2-1) in LUAD (mean \u0026Delta;PSI of -0.3). NKX2-1 has two known isoforms[\u003cspan class=\"CitationRef\"\u003e41\u003c/span\u003e], where the PTSE \u003cem\u003eNKX2-1_27289_ES\u003c/em\u003e leads to a shorter protein isoform. Because the mean \u0026Delta;PSI value of NKX2-1_27289_ES is negative, the shorter isoform NKX2-1 potentially shows higher prevalence in cancer samples than in normal samples. NKX2-1 is a lineage TF for lung[\u003cspan class=\"CitationRef\"\u003e42\u003c/span\u003e]. Functional differences between the two isoforms have also been shown[\u003cspan class=\"CitationRef\"\u003e43\u003c/span\u003e]. Because the minimum Chronos score of NKX2-1 is -1.4 (Supplementary Table \u003cspan class=\"InternalRef\"\u003eS7\u003c/span\u003e), indicating its essentiality in cancer proliferation, perhaps it\u0026apos;s the higher prevalence of the shorter isoform that determines the cancer dependency of NKX2-1. Overall, this analysis identified TFs that are not only essential for cancer proliferation, but also showed significant differences between normal and paired cancer samples with respect to AS events. This could imply that cancer dependency of these TFs are reliant on the corresponding TF isoforms.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eMany AS events showing significant differences between paired normal and cancer samples were found across 14 cancer types, affecting nearly 50% of known TFs with sufficient annotations. Several of these TFs were found to be common in multiple cancer types, and many others were uniquely affected in a cancer type. These results highlight a pancancer significance of aberrant TF splicing in cancer. We found evidence of the impact of cancer-specific AS events on the DNA-binding and gene regulatory (activating/repressing) abilities of TFs. Further evidence for the functional impact of cancer-specific AS events on TFs were found by exploring whether the affected TFs were essential for cancer cell line proliferation. Overall, these findings could be used by future research to understand AS-driven cancer molecular mechanisms, which could guide identifications of novel therapeutic targets.\u003c/p\u003e\u003cp\u003eDifferent algorithms to detect and quantify AS events have been proposed in the literature, e.g., SpliceSeq[\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e], Sp1Adder[\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e], and Cufflinks[\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. All of the findings in this work are based on the TCGA SpliceSeq database, which used the SpliceSeq algorithm to quantify AS events. Future work could evaluate whether these findings are robust across different algorithms.\u003c/p\u003e\u003cp\u003eGene regulatory associations based on gene level expressions of TFs have provided interesting insights in cancer[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Our results show that there is a large-scale potential functional impact of cancer-specific AS events on TFs, which could perturb gene regulatory associations. Future inference of gene regulatory associations in cancer could benefit from considering expressions of TFs not just on the gene level but also on the isoform level.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch3\u003eAuthor contributions\u003c/h3\u003e\n\u003cp\u003eKN and JB conceived the study. KN performed all of the work. KN, OT, and JB analyzed the results. JB provided the computational infrastructure. KN, OT, and JB wrote the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003ch3\u003eCompeting interests\u003c/h3\u003e\n\u003cp\u003eThe authors claim no competing interest.\u003c/p\u003e\n\u003ch3\u003eData and code availability\u003c/h3\u003e\n\u003cp\u003eAll data were taken from public databases and the relevant links have been provided within the manuscript. The code to reproduce the work is publicly available at https://doi.org/10.5281/zenodo.17279821. The data extracted from the public databases for reproducing this work is available at https://doi.org/10.5281/zenodo.17250324.\u003c/p\u003e\n\u003ch3\u003eFunding\u003c/h3\u003e\n\u003cp\u003eThis work was supported by the Excellence Strategy of the Federal Government of Germany and the L\u0026auml;nder, the Deutsche Forschungsgemeinschaft (grant number 517063424), and the Bundesministerium f\u0026uuml;r Bildung und Forschung (grant numbers 01ZX1908A, 01ZX2208A, and 031L0287B).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003ePan Q, Shai O, Lee LJ, Frey BJ, Blencowe BJ. Deep surveying of alternative splicing complexity in the human transcriptome by high-throughput sequencing. Nat Genet. 2008;40(12):1413\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBaralle FE, Giudice J. Alternative splicing as a regulator of development and tissue identity. Nat Rev Mol Cell Biol. 2017;18(7):437\u0026ndash;51.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eScotti MM, Swanson MS. RNA mis-splicing in disease. Nat Rev Genet. 2016;17(1):19\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eStanley RF, Abdel-Wahab O. Dysregulation and therapeutic targeting of RNA splicing in cancer. Nat Cancer. 2022;3(5):536\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRyan M, Wong WC, Brown R, Akbani R, Su X, Broom B, Melott J, Weinstein J. TCGASpliceSeq a compendium of alternative mRNA splicing in cancer. Nucleic Acids Res. 2016;44(D1):D1018\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSveen A, Kilpinen S, Ruusulehto A, Lothe RA, Skotheim RI. Aberrant RNA splicing in cancer; expression changes and driver mutations of splicing factor genes. Oncogene. 2016;35(19):2413\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi Y, Sahni N, Pancsa R, McGrail DJ, Xu J, Hua X, Coulombe-Huntington J, Ryan M, Tychhon B, Sudhakar D, Hu L, Tyers M, Jiang X, Lin SY, Babu MM, Yi S. Revealing the determinants of widespread alternative splicing perturbation in cancer. Cell Rep. 2017;21(3):798\u0026ndash;812.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi J, Pan T, Chen L, Wang Q, Chang Z, Zhou W, Li X, Xu G, Li X, Li Y, Zhang Y. Alternative splicing perturbation landscape identifies RNA binding proteins as potential therapeutic targets in cancer. Mol Ther Nucleic Acids. 2021;24:792\u0026ndash;806.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng R, Xiao L, Zhou W, Jin X, Xu Z, Xu C, Wang P, Luo M, Wang M, Ma K, Cao H, Huang Y, Lin X, Pang F, Li Y, Jiang Q. A pan-cancer analysis of alternative splicing of splicing factors in 6904 patients. Oncogene. 2021;40(35):5441\u0026ndash;50.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBelluti S, Rigillo G, Imbriano C. Transcription factors in cancer: When alternative splicing determines opposite cell fates. Cells. 2020;9(3):760.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLambert SA, Jolma A, Campitelli LF, Das PK, Yin Y, Albu M, Chen X, Taipale J, Hughes TR, Weirauch MT. The human transcription factors. Cell. 2018;172(4):650\u0026ndash;65.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBradner JE, Hnisz D, Young RA. Transcriptional addiction in cancer. Cell. 2017;168(4):629\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBushweller JH. Targeting transcription factors in cancer - from undruggable to reality. Nat Rev Cancer. 2019;19(11):611\u0026ndash;24.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDarnell JE Jr. Transcription factors as targets for cancer therapy. Nat Rev Cancer. 2002;2(10):740\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLambourne L, Mattioli K, Santoso C, Sheynkman G, Inukai S, Kaundal B, Berenson A, Spirohn-Fitzgerald K, Bhattacharjee A, Rothman E, Shrestha S, Laval F, Carroll BS, Plassmeyer SP, Emenecker RJ, Yang Z, Bisht D, Sewell JA, Li G, Prasad A, Phanor S, Lane R, Moyer DC, Hunt T, Balcha D, Gebbia M, Twizere JC, Hao T, Holehouse AS, Frankish A, Riback JA, Salomonis N, Calderwood MA, Hill DE, Sahni N, Vidal M, Bulyk ML, Fuxman Bass JI. Widespread variation in molecular interactions and regulatory properties among transcription factor isoforms. Mol Cell. 2025;85(7):1445\u0026ndash;66.e13.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFan YC, Min L, Chen H, Liu YL. Alternative splicing isoform of T cell factor 4K suppresses the proliferation and metastasis of non-small cell lung cancer cells. Genet Mol Res. 2015;14(4):14009\u0026ndash;18.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTsedensodnom O, Koga H, Rosenberg SA, Nambotin SB, Carroll JJ, Wands JR, Kim M. Identification of T-cell factor-4 isoforms that contribute to the malignant phenotype of hepatocellular carcinoma cells. Exp Cell Res. 2011;317(7):920\u0026ndash;31.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang Y, Qian J, Gu C, Yang Y. Alternative splicing and cancer: a systematic review. Signal Transduct Target Ther. 2021;6(1):78.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCorces MR, Granja JM, Shams S, Louie BH, Seoane JA, Zhou W, Silva TC, Groeneveld C, Wong CK, Cho SW, Satpathy AT, Mumbach MR, Hoadley KA, Robertson AG, Sheffield NC, Felau I, Castro MAA, Berman BP, Staudt LM, Zenklusen JC, Laird PW, Curtis C, Cancer Genome Atlas Analysis Network, Greenleaf WJ, Chang HY. The chromatin accessibility landscape of primary human cancers. Science. 2018;362(6413):eaav1898.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eArafeh R, Shibue T, Dempster JM, Hahn WC, Vazquez F. The present and future of the Cancer Dependency Map. Nat Rev Cancer. 2025;25(1):59\u0026ndash;73.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCancer Genome Atlas Research Network, Weinstein JN, Collisson EA, Mills GB, Shaw KRM, Ozenberger BA, Ellrott K, Shmulevich I, Sander C, Stuart JM. The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet. 2013;45(10):1113\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSigrist CJA, de Castro E, Cerutti L, Cuche BA, Hulo N, Bridge A, Bougueleret L, Xenarios I. New and continuing developments at PROSITE. Nucleic Acids Res. 2013;41(Database issue):D344\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, Tosatto SCE, Paladin L, Raj S, Richardson LJ, Finn RD, Bateman A. Pfam: The protein families database in 2021. Nucleic Acids Res. 2021;49(D1):D412\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLetunic I, Khedkar S, Bork P. SMART: recent updates, new developments and status in 2020. Nucleic Acids Res. 2021;49(D1):D458\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSoto LF, Li Z, Santoso CS, Berenson A, Ho I, Shen VX, Yuan S, Fuxman Bass JI. Compendium of human transcription factor effector domains. Mol Cell. 2022;82(3):514\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang Y, Yao X, Zhou H, Wu X, Tian J, Zeng J, Yan L, Duan C, Liu H, Li H, Chen K, Hu Z, Ye Z, Xu H. OncoSplicing: an updated database for clinically relevant alternative splicing in 33 human cancers. Nucleic Acids Res. 2022;50(D1):D1340\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCai Z, Qian ZY, Jiang H, Ma N, Li Z, Liu LY, Ren XX, Shang YR, Wang JJ, Li JJ, Liu DP, Zhang XP, Feng D, Ni QZ, Feng YY, Li N, Zhou XY, Wang X, Bao Y, Zhang XL, Deng YZ, Xie D. HPCL3s promotes hepatocellular carcinoma metastasis by activating β-catenin signaling. Cancer Res. 2018;78(10):2536\u0026ndash;49.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu Y, Liu W, Jiang X, Li J, Liu Q, Su F, Ruan S, Zhang Z, Tao F. Regulation of hPCL3 isoforms\u0026rsquo; ubiquitination by TRIM21 in non-small cell lung cancer progression. Life Sci Alliance [Internet]. 2023;6(10). Available from: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.26508/lsa.202302060\u003c/span\u003e\u003cspan address=\"10.26508/lsa.202302060\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAbdelfettah S, Boulay G, Dubuissez M, Spruyt N, Garcia SP, Rengarajan S, Loison I, Leroy X, Rivera MN, Leprince D. hPCL3S promotes proliferation and migration of androgen-independent prostate cancer cells. Oncotarget. 2020;11(12):1051\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang S, Robertson GP, Zhu J. A novel human homologue of Drosophila polycomblike gene is up-regulated in multiple cancers. Gene. 2004;343(1):69\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCox CJ, Espinoza HM, McWilliams B, Chappell K, Morton L, Hjalt TA, Semina EV, Amendt BA. Differential regulation of gene expression by PITX2 isoforms. J Biol Chem. 2002;277(28):25001\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGrassmeyer J, Mukherjee M, deRiso J, Hettinger C, Bailey M, Sinha S, Visvader JE, Zhao H, Fogarty E, Surendran K. Elf5 is a principal cell lineage specific transcription factor in the kidney that contributes to Aqp2 and Avpr2 gene expression. Dev Biol. 2017;424(1):77\u0026ndash;89.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLu Y, Berenson A, Lane R, Guelin I, Li Z, Chen Y, Shah S, Yin M, Soto-Ugaldi LF, Fiszbein A, Fuxman Bass JI. A large-scale cancer-specific protein-DNA interaction network. Life Sci Alliance. 2024;7(10):e202402641.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZaika AI, Slade N, Erster SH, Sansome C, Joseph TW, Pearl M, Chalas E, Moll UM. DeltaNp73, a dominant-negative inhibitor of wild-type p53 and TAp73, is up-regulated in human tumors. J Exp Med. 2002;196(6):765\u0026ndash;80.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFillippovich I, Sorokina N, Gatei M, Haupt Y, Hobson K, Moallem E, Spring K, Mould M, McGuckin MA, Lavin MF, Khanna KK. Transactivation-deficient p73alpha (p73Deltaexon2) inhibits apoptosis and competes with p53. Oncogene. 2001;20(4):514\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBaek S, Goldstein I, Hager GL. Bivariate genomic footprinting detects changes in transcription factor activity. Cell Rep. 2017;19(8):1710\u0026ndash;22.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHe HH, Meyer CA, Chen MW, Jordan VC, Brown M, Liu XS. Differential DNase I hypersensitivity reveals factor-dependent chromatin dynamics. Genome Res. 2012;22(6):1015\u0026ndash;25.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKeshamouni VG, Reddy RC, Arenberg DA, Joel B, Thannickal VJ, Kalemkerian GP, Standiford TJ. Peroxisome proliferator-activated receptor-gamma activation inhibits tumor progression in non-small-cell lung cancer. Oncogene. 2004;23(1):100\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChang TH, Szabo E. Induction of Differentiation and Apoptosis by Ligands of Peroxisome Proliferator-activated Receptor γ in Non-Small Cell Lung Cancer. Cancer Res. 2000;60(4):1129\u0026ndash;38.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBehan FM, Iorio F, Picco G, Gon\u0026ccedil;alves E, Beaver CM, Migliardi G, Santos R, Rao Y, Sassi F, Pinnelli M, Ansari R, Harper S, Jackson DA, McRae R, Pooley R, Wilkinson P, van der Meer D, Dow D, Buser-Doepner C, Bertotti A, Trusolino L, Stronach EA, Saez-Rodriguez J, Yusa K, Garnett MJ. Prioritization of cancer therapeutic targets using CRISPR-Cas9 screens. Nature. 2019;568(7753):511\u0026ndash;6.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHamdan H, Liu H, Li C, Jones C, Lee M, deLemos R, Minoo P. Structure of the human Nkx2.1 gene. Biochim Biophys Acta. 1998;1396(3):336\u0026ndash;48.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKimura S, Hara Y, Pineau T, Fernandez-Salguero P, Fox CH, Ward JM, Gonzalez FJ. The T/ebp null mouse: thyroid-specific enhancer-binding protein is essential for the organogenesis of the thyroid, lung, ventral forebrain, and pituitary. Genes Dev. 1996;10(1):60\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLi C, Cai J, Pan Q, Minoo P. Two functionally distinct forms of NKX2.1 protein are expressed in the pulmonary epithelium. Biochem Biophys Res Commun. 2000;270(2):462\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRyan MC, Cleland J, Kim R, Wong WC, Weinstein JN. SpliceSeq: a resource for analysis and visualization of RNA-Seq data on alternative splicing and its functional impacts. Bioinformatics. 2012;28(18):2385\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKahles A, Ong CS, Zhong Y, R\u0026auml;tsch G. SplAdder: identification, quantification and testing of alternative splicing events from RNA-Seq data. Bioinformatics. 2016;32(12):1840\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTrapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, Pimentel H, Salzberg SL, Rinn JL, Pachter L. Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks. Nat Protoc. 2012;7(3):562\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBen Guebila M, Lopes-Ramos CM, Weighill D, Sonawane AR, Burkholz R, Shamsaei B, Platig J, Glass K, Kuijjer ML, Quackenbush J. GRAND: a database of gene regulatory network models across human conditions. Nucleic Acids Res. 2022;50(D1):D610\u0026ndash;21.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"oncogene","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"onc","sideBox":"Learn more about [Oncogene](http://www.nature.com/onc/)","snPcode":"41388","submissionUrl":"https://mts-onc.nature.com/cgi-bin/main.plex","title":"Oncogene","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7799323/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7799323/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eCancer shows aberrant alternative splicing (AS), which could functionally perturb proteins required for normal cellular behavior. Transcription factors (TFs) are proteins that regulate gene transcription. Some evidence about AS-driven perturbation of TFs in cancer has been documented. A comprehensive systematic analysis of how cancer-specific AS could affect functions of TFs is missing. Such an analysis could reveal the molecular mechanisms of cancer progression due to transcription misregulations, and thus could help identify therapeutic targets. Here, we systematically analyzed potential functional perturbations of TFs due to AS across 15 cancer types. By analyzing AS patterns, we identified 2 118 perturbed AS events (i.e., events showing significant AS pattern differences between normal and paired cancer samples) that affect 718 TFs across 14 cancer types. In 189 TFs, the perturbed AS events affected known functional domains of TFs. Indirect evidence for the functional impact of cancer-specific AS on TFs were also found: First, by relating AS patterns of the perturbed AS events with the DNA-binding and regulatory activity of TFs. Second, by using cancer dependency data to explore whether the affected TFs are essential for cancer cell line proliferation. Our findings show a large-scale functional perturbation of TFs due to cancer-specific AS.\u003c/p\u003e","manuscriptTitle":"Aberrant splicing in human cancer has a large-scale functional impact on transcription factors","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-11 17:01:44","doi":"10.21203/rs.3.rs-7799323/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"revise","date":"2026-02-04T14:07:07+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-01-30T16:57:49+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-01-10T22:05:03+00:00","index":2,"fulltext":"This content is not available."},{"type":"editorInvitedReview","content":"This content is not available.","date":"2025-11-15T20:13:45+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2025-10-31T14:43:12+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2025-10-31T12:03:28+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-10-08T11:20:56+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-10-07T12:00:47+00:00","index":"","fulltext":""},{"type":"submitted","content":"Oncogene","date":"2025-10-07T12:00:46+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"oncogene","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"onc","sideBox":"Learn more about [Oncogene](http://www.nature.com/onc/)","snPcode":"41388","submissionUrl":"https://mts-onc.nature.com/cgi-bin/main.plex","title":"Oncogene","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"19ab784b-f0f9-4bc0-bc8d-caa726e01a4a","owner":[],"postedDate":"November 11th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":57224321,"name":"Biological sciences/Cancer"},{"id":57224322,"name":"Biological sciences/Molecular biology/Transcriptomics"}],"tags":[],"updatedAt":"2026-04-03T19:35:41+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-11 17:01:44","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7799323","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7799323","identity":"rs-7799323","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}