{"paper_id":"3a71562a-7caf-4571-9ebf-b361cc5b90c2","body_text":"1 \n \nChallenges and opportunities for drug repurposing in cancers based on 1 \nsynthetic lethality induced by tumor suppressor gene mutations  2 \nMichael Vermeulen1, Andrew W. Craig3,4, Tomas Babak1,2 3 \n1. Department of Biology, Queen’s University, Kingston, Ontario, Canada 4 \n2. Leapfrog Bio, San Mateo, CA, USA 5 \n3. Department of Biomedical and Molecular Sciences, Queen’s University, Kingston, Ontario, Canada 6 \n4. Cancer Biology and Genetics Division, Sinclair Cancer Research Institute, Kingston, Ontario, Canada 7 \n 8 \nAbstract 9 \nAlthough two-thirds of cancers arise from loss-of-function mutations in tumor suppressor genes, there 10 \nare few approved targeted therapies linked to these alterations. Synthetic lethality offers a promising 11 \nstrategy to treat such cancers by targeting vulnerabilities unique to cancer cells with these mutations. 12 \nTo identify clinically relevant synthetic lethal interactions, we analyzed genome-wide CRISPR/Cas9 13 \nknock-out (KO) viability screens from the Cancer Dependency Map and evaluated their clinical 14 \nrelevance in patient tumors through mutual exclusivity, a pattern indicative of synthetic lethality. Indeed, 15 \nwe found significant enrichment of mutual exclusivity for interactions involving cancer driver genes 16 \ncompared to non-driver mutations. To identify therapeutic opportunities, we integrated drug sensitivity 17 \ndata to identify inhibitors that mimic the effects of CRISPR-mediated KO. This approach revealed 18 \npotential drug repurposing opportunities, including BRD2 inhibitors for bladder cancers with ARID1A 19 \nmutations and SIN3A-mutated cell lines showing sensitivity to nicotinamide phosphoribosyltransferase 20 \n(NAMPT) inhibitors. However, we discovered that pharmacological inhibitors often fail to phenocopy KO 21 \nof matched drug targets, with only a small fraction of drugs inducing similar effects. This discrepancy 22 \nreveals fundamental differences between pharmacological and genetic perturbations, emphasizing the 23 \nneed for approaches that directly assess the interplay of loss-of-function mutations and drug activity in 24 \ncancer models. 25 \n 26 \n 27 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n2 \n \nAuthor Summary 28 \nSynthetic lethality is an emerging approach for targeting a biological dependency in cancer cells that 29 \ndoes not harm normal cells. This strategy is particularly valuable for targeting loss-of-function mutations 30 \nin tumor suppressor genes, which are more challenging to directly target. In an effort to accelerate 31 \ntreatments for cancer patients, we aimed to map out these dependencies and overlap them with 32 \nresponses to available drugs. We discovered different outcomes when a protein is targeted by a drug 33 \nversus when that same target is disrupted genetically. Thus, if a drug is to be effectively repurposed as 34 \nsynthetic lethal agent, feasibility studies must capture drug biology, ideally by test the drug empirically 35 \nin relevant cancer models. A second notable discovery is that in vitro synthetic lethal interactions 36 \ninvolving cancer driver genes are significantly more likely to exhibit consistent patterns, such as mutual 37 \nexclusivity in human tumor samples. This is important since selection of relevant cell lines is often 38 \ncritical in drug development to maximize potential for translation to clinical responses.  39 \n 40 \n1 Introduction 41 \nSynthetic lethality (SL) provides a compelling framework for developing cancer therapies tailored to the 42 \nspecific genetic profiles of individual patient tumors. SL relies on the genetic interaction between two 43 \ngenes where disruption of either gene alone is tolerated, but their simultaneous disruption leads to cell 44 \ndeath. Applying this strategy in oncology usually involves pharmacologically targeting a protein whose 45 \ninhibition is lethal to cancer cells harboring a specific loss-of-function (LoF) mutation in a partnering 46 \ngene, but spare healthy cells without these mutations (S1A Fig). The appeal of exploiting SL lies in the 47 \npotential to increase treatment efficacy, minimize adverse effects and expand the range of actionable 48 \ntherapeutic targets, particularly tumor suppressor genes (TSGs) that currently lack effective therapies 49 \n[1,2]. An exemplar of this approach is the use of PARP inhibitors, which harness SL to exploit DNA 50 \nrepair deficiencies in homologous recombination deficient (HRD) cancers, commonly driven by LoF 51 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n3 \n \nalterations in BRCA1 and BRCA2 [3]. Generating over $3 billion annually, PARP inhibitors have 52 \nbecome standard of care treatment for several HRD cancers, and are increasingly used in combination 53 \ntherapies that prove the value of exploiting the potential of SL-driven oncology. 54 \nDespite the success of PARP inhibitors, broader clinical adoption of additional SL drugs has not been 55 \nas effective as predicted and is limited to a handful of clinical-stage developments, none of which have 56 \nyet been approved. However, progress may have been limited by a lack of systematic SL mapping 57 \nabilities in human cells. Recent high-throughput screening efforts hold promise for expanding SL 58 \ntreatments, allowing unbiased target exploration across the genome. Genome-wide CRISPR and RNA 59 \ninterference-based genetic screens, including those from the Cancer Dependency Map Project 60 \n(DepMap), OnTarget, and Project DRIVE, have substantially expanded the repertoire of potential 61 \ntherapeutic targets.[4] Since it’s public release in 2018, DepMap has become an essential tool for 62 \nhypothesis generation, facilitating the translation of numerous SL interactions into clinical trials. 63 \nOngoing trials include targeting PRMT5 (NCT05732831; Tango Therapeutics), and MAT2A in MTAP-64 \ndeleted cancers (NCT04794699; IDEAYA Biosciences), and WRN in MSI or mismatch repair deficient 65 \ntumors (NCT05838768; Novartis). As novel actionable SL targets and corresponding therapeutic 66 \nagents are identified, the potential of precision oncology advances toward increasingly meaningful 67 \nclinical impact.  68 \nNumerous studies have developed methods for identifying additional SL targets within genetically 69 \ndefined contexts. Approaches range from univariate tests [5–8], linear models such as mixed-model 70 \nregressions [9], ANOVA [10], and machine learning techniques including random forests (RF) [11–16], 71 \nneural networks [17] and autoencoders [18,19]. Among these, RF models have gained prominence for 72 \ntheir capacity to identify molecular biomarkers linked to gene essentiality and drug sensitivity in cancer 73 \ncell lines. RF models capture complex, multivariate, and non-linear relationships that univariate tests 74 \nand linear models often fail to detect. RFs can account for interactions between variables, offer 75 \ninterpretable feature importance rankings, and provide insights into complex biological systems. 76 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n4 \n \nAdditionally, they are computationally efficient and adept at integrating diverse omics data alongside 77 \nconfounding variables, making them especially well-suited for analyzing high-dimensional datasets, 78 \nsuch as those produced by the DepMap project [11].  79 \nDespite methodological advances, statistical models often yield thousands of potential SL interactions, 80 \nnecessitating the use of additional filters to enrich for the most clinically actionable interactions. To 81 \nachieve this, previous studies have leveraged biological pathway data [7,20], protein-protein 82 \ninteractions [9,10], drug sensitivity profiles [5,20] and clinical information from The Cancer Genome 83 \nAtlas (TCGA) such as tumor gene expression [5], survival statistics [5] and mutual exclusivity [21,22]. 84 \nMutual exclusivity is a powerful readout for filtering SL interactions using patient tumor mutations [23]. 85 \nThe concept is based on the observation that certain genetic alterations rarely co-occur in tumors, 86 \nimplying negative selective pressure on their coexistence (S1B Fig). Mutual exclusivity is thus a clinical 87 \nindicator for SL. Focusing on gene pairs or pathways that exhibit mutual exclusivity in tumors can 88 \nenrich potential in vitro SL interactions that are more likely to translate into therapeutic benefit.  89 \nRecognizing that drugs can elicit complex pharmacological responses on cell biology, several initiatives 90 \nhave systematically screened drugs across extensive cancer cell line panels. Projects such as PRISM 91 \n(Profiling Relative Inhibition Simultaneously in Mixtures) [12], GDSC (Genomics of Drug Sensitivity in 92 \nCancer) [24], and CTD2 (Cancer Target Discovery and Development) [5] have significantly advanced 93 \nour understanding of drug response across a diverse range of cancer types. These datasets enable the 94 \nmodelling of relationships between molecular features and drug response [9,12,25], providing valuable 95 \ninsight into mechanisms of action (MoA) as well as biomarkers of drug sensitivity and resistance 96 \n[26,27]. Integrating drug sensitivity profiles with SL interactions can highlight cases where 97 \npharmacological inhibition and genetic perturbation of the same gene/protein induces similar synthetic 98 \nlethal effects, thus supporting the reliability of SL interactions.  99 \nIn pursuit of robust pharmacogenomic interactions and potential drug repositioning opportunities, two 100 \nprevious studies have conducted comprehensive comparisons of drug and CRISPR viability. Gonçalves 101 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n5 \n \net al integrated drug sensitivity profiles from the GDSC with DepMap gene viability profiles, finding that 102 \napproximately 25% of 376 targeted oncology drugs had effects consistent with CRISPR deletion of the 103 \nsame targets [9]. Similarly, using PRISM drug sensitivity profiles, Corsello et al. [12] showed that 15% 104 \nof active targeted oncology drugs and 0.8% of active non-oncology drugs displayed sensitivity profiles 105 \nresembling CRISPR KO of their intended targets. These findings suggest that, while pharmacological 106 \ninhibition and genetic disabling of a gene can produce similar outcomes, it is more common that the two 107 \nmodes of targeting a gene product differ, which violates a fundamental assumption in SL studies. 108 \nRecognizing that there are typically differences between pharmacological inhibition and genetic KO of a 109 \ntarget is an important step in assessing the potential of drug repositioning, such as incorporating more 110 \nelements of drug biology.  111 \nIn this study, we evaluate the principles underlying SL interactions identified from DepMap pooled 112 \nCRISPR KO screens to assess their potential as druggable opportunities, with a specific focus on 113 \nuncovering novel genetically targeted applications for existing clinical grade drugs. In contrast with 114 \nearlier studies, we found that only a small fraction of SL interactions identified in CRISPR KO screens 115 \nare likely to be replicated in tumors, with a significantly higher probability for interactions involving driver 116 \ngenes. For the 6,550 drugs tested in cell viability screens, concordance with genetic KO was 117 \nunexpectedly low, highlighting the need for repurposing approaches that effectively capture drug 118 \nbiology, such as empirical pharmacogenomic screens or additional mechanistic insights. 119 \n 120 \n2 Results 121 \nConstructing a driver gene-centric network for repurposing and targeting cancer vulnerabilities 122 \nGenetic dependencies network 123 \nOur approach aimed to identify mutation-specific vulnerabilities in cancer cell lines by stratifying 124 \nfrequently mutated genes based on mutation status and assessing differential fitness effects across 125 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n6 \n \ngenome-wide gene KO’s. Systematically evaluating each mutated gene against all potential KO’s 126 \nenabled us to build a network of genetic dependencies central to our analyses. We employed an 127 \nadaptation of the Pan-Cancer Inferred Synthetic Lethalities (PARIS) framework [13] which applies the 128 \nBoruta algorithm – a Random Forest-based feature selection method designed to capture all significant 129 \npredictive features. Through our modified framework, we identified mutations associated with enhanced 130 \nsensitivity or resistance to gene inhibition, systematically mapping critical genetic dependencies across 131 \ndiverse cancer cell line models without the need to exhaustively test each gene pair. This approach 132 \nenables a precise and scalable characterization of mutation-driven dependencies, enabling us to 133 \nexplore potential therapeutic targets and drug repurposing opportunities. 134 \nOur analysis utilized input data from DepMap (release 23Q2) which included detailed mutation profiles 135 \nand gene essentiality scores obtained from CRISPR LoF (CRISPRi) screens across 1,087 cancer cell 136 \nlines. Gene essentiality scores – which reflect the impact of gene KO’s on cell proliferation – served as 137 \nresponse variables. The mutation data, including both hotspot and damaging mutations with detailed 138 \nzygosity information were used as predictor variables. We performed this analysis across all cell lines 139 \nwith complete mutation and essentiality profiles, identifying interactions both in a pan-cancer context 140 \nand within 13 specific cancer types. After adjusting for confounders including microsatellite instability 141 \nand cancer type using linear regression (FDR < 0.1) and excluding interactions with insufficient 142 \nmutation data (< 4 mutations), our final pan-cancer network contained 1,428 significant genetic 143 \ninteractions (details provided in Methods). To clarify terminology, within an SL interaction, we define 144 \nthe mutated driver gene in the cell line as the ‘source’ gene and the gene targeted by CRISPR KO as 145 \nthe ‘target’ gene. Comprehensive details of all interactions, both pan-cancer and cancer-type specific, 146 \nare documented in Table S1 and Table S2.  147 \nIdentification of clinically relevant SL interactions 148 \nSince the source data was generated in cancer cell lines, which rarely recapitulate the complexity of 149 \nprimary tumors [28–30], we were interested in exploring parameters that predicted which SL 150 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n7 \n \ninteractions are more likely to translate into clinical benefit on the basis of tumor biology. We reasoned 151 \nthat genetic interaction with driver genes [31–33], which are mutated at epidemiologically 152 \noverrepresented levels in patients with cancer, may be more likely to conserved since cell lines tend to 153 \nbe dependent on their driver mutations [34]. As these genes are also more frequently mutated in patient 154 \ntumors, SL interactions could reveal treatment opportunities with significant potential for patient benefit. 155 \nTo evaluate whether driver gene SL interactions are more likely to be conserved in tumors, we 156 \nanalyzed mutual exclusivity (ME) using TCGA tumor mutation and copy number alteration data (Fig 1). 157 \nME is based on the observation that certain genetic alterations co-occur less frequently than expected 158 \nin patient tumors, suggesting an intrinsic incompatibility and potential SL relationship (Fig S1B). As a 159 \nclinical indicator for SL, ME allows us to focus on gene pairs with greater translational potential. 160 \nAccordingly, for each cancer type we applied hypergeometric tests to assess ME among the statistically 161 \npowered gene interactions within our network (Methods). As hypothesized, in vitro SL interactions 162 \ninvolving driver mutations demonstrated significantly stronger ME signals in human tumors (P < 0.05) 163 \ncompared to those lacking drivers (Fig 2C), even when controlling for power differences arising from 164 \nhigher mutation rates of drivers versus non-drivers (Fig S2). This observation supports the central role 165 \nof driver mutations in dictating mutual exclusivity patterns and confirms that in vitro SL interactions 166 \ncontaining drivers are biologically relevant in clinical settings.  167 \nNetwork visualization of the genetic interactions revealed central hubs enriched by driver genes, such 168 \nas TP53, APC, RB1 and Ras/Raf/MAPK pathway oncogenes (Fig 2A). Given the clustering of central 169 \nhubs around common driver genes and the stronger ME signals observed in driver interactions relative 170 \nto non-driver interactions, we refined the network to focus exclusively on interactions involving ‘source’ 171 \ndriver mutations - mutations pre-existing in the genomic profiles of the cell lines. These are interactions 172 \ninvolving a driver mutation inherent to a cancer cell line. This subset confined the network to 510 gene 173 \npairs across 448 unique genes, including 46 driver genes (Fig 2B). We prioritized this subset to reduce 174 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n8 \n \nthe search space and to concentrate on interactions more likely to have clinical relevance and 175 \ntranslational potential.  176 \nTo improve interpretability, we introduced directionality to the genetic interactions using the correlation 177 \ncoefficient; negative values indicated increased sensitivity in the mutated state, while positive values 178 \nindicated sensitivity in the wild-type. For drug repurposing purposes we were interested in SL 179 \ninteractions, specifically cases where mutated cell lines showed heightened sensitivity to partner gene 180 \nKO. 181 \nSignificant driver genetic interactions in the pan-cancer network were depicted in a heatmap (FDR < 182 \n0.1; Fig 3A), where numerous established dependencies were evident. These included self-paired 183 \noncogenes such as BRAF (Box 1) and PIK3CA (Box 4), paralog dependencies such as SMARCA2-184 \nSMARCA4 (Fig 3C; Box 3), STAG1-STAG2 (Fig 3C; Box 2), RPL22-RP22L1, PDS5A-PDS5B, and 185 \nARID1A-ARID1B, and other established pathway dependencies such as SHOC2-NRAS, PTPN11-186 \nBRAF, CREBBP-EP300 and RB1-CDK6 (Box 3). The enrichment of well-known, and experimentally 187 \nconfirmed genetic dependencies gave us confidence that we were capturing biologically relevant signal 188 \nwith power to identify clinically relevant cases downstream in our effort to repurpose cancer therapies 189 \nexploiting SL interactions. Our analysis revealed extensive SL and resistance interactions (Fig S3).    190 \nOverlap with PRISM 191 \nAfter establishing our driver-focused genetic interaction network, we integrated 6,550 drug sensitivity 192 \nprofiles (6,337 unique drugs) from the PRISM 23Q4 dataset, encompassing two distinct screening 193 \nefforts: Repurposing-Primary (n = 5362) and recently released Repurposing-1M (n = 1278). Drug-gene 194 \nassociations were mapped based on putative gene targets derived from Citeline and PRISM drug target 195 \ndatabases. Among the pan-cancer interactions we established, 20.5% (104/510) had at least one drug 196 \nwith a matching target in the PRISM dataset. To explore the behavior of drug profiles targeting genes 197 \nimplicated in our genetic interactions, we implemented two analytical strategies: (i) examining the linear 198 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n9 \n \nrelationship between drug sensitivity and CRISPR KO effect scores, and (ii) assessing drug sensitivities 199 \nin cell lines stratified by driver mutation status (Fig 1). Both analyses – adjusted for microsatellite 200 \ninstability (MSI), cell growth rates, and cancer types – aimed to identify interaction pairs where the drug 201 \neffect on cell proliferation mirrored that of gene KO. While the mutation-stratified analysis is limited by 202 \nsmaller sample sizes, particularly in cancer-specific subsets and for less common driver mutations, it 203 \nprovides direct insights into context-specific therapeutic potential with translational value. In contrast, 204 \nthe linear correlation approach leverages the entirety of the dataset, providing better statistical power to 205 \ndetect strong and generalizable dependencies. 206 \nWe initially assessed correlations between drug sensitivity and gene essentiality for druggable targets 207 \nin pan-cancer genetic interaction pairs. Among these targets, 40.3% (42/104) displayed a significant 208 \nlinear correlation between drug sensitivity and CRISPR KO (FDR < 0.05; S3 Table), indicating that 209 \nthese genetic dependencies may be pharmacologically targeted to induce SL. Notably, many of these 210 \n'druggable SL' pairs involved well-characterized cancer genes, such as MDM2, TP53, MCL1, BRAF, 211 \nPTPN11, PIK3CA, EGFR, BCL2L1, CDK6, and ERBB2, for which targeted therapies are already 212 \navailable, serving as positive controls. For example, dependencies on MDM2 and TP53 strongly 213 \ncorrelated with sensitivity to various MDM2 inhibitors, underscoring the parallel effects of 214 \npharmacological inhibition and gene KO on MDM2-mediated p53 repression. Other key correlations 215 \nwere observed with cell cycle-related genes, such as MCL1 dependencies with MCL1 inhibitors 216 \n(AZD5991, AMG-176), CDK6 with CDK4/6 inhibitors (Ribociclib, Palbociclib, Trilaciclib), and BCL2L1 217 \nwith BCL inhibitors (Navitoclax, ABT-737; Fig S4). Furthermore, dependencies in epigenetic 218 \nregulators—such as EP300 with the p300/CBP inhibitor A-485 and BRD2 with BET inhibitors including 219 \nOTX015, TEN-010, and CPI-203 – highlight the potential for therapeutic intervention in these drug-gene 220 \nrelationships (Fig S4). 221 \nHowever, across the entire PRISM database, correlations between drug sensitivity and CRISPR KO 222 \nessentiality of drug targets were rare. Of all PRISM drugs, only 154 (3%) displayed a significant 223 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n10 \n \ncorrelation with any target gene essentiality profile (FDR < 0.1; Table S3). Additionally, for each drug, 224 \nwe compared the correlation coefficient between drug sensitivity and effect scores for all putative target 225 \ngenes against a distribution of randomly permuted effect scores (Fig 4). In the majority of cases, 226 \nobserved correlations did not significantly deviate from the permuted distribution, with the exception of 227 \na skewed tail of highly correlated pairs, mostly from highly targeted oncology drugs.  228 \nThese findings suggest that while CRISPRi screens are effective at identifying essential genes, the 229 \ngenetic dependencies they reveal may not directly correlate with drug efficacy. Small-molecule 230 \ninhibitors typically target only the catalytic function of proteins, which can activate compensatory cellular 231 \nmechanisms, context-dependent dependencies, and off-target effects, complicating the translation of 232 \ngenetic vulnerabilities into therapeutics. Furthermore, the promiscuity of many drugs, with multiple 233 \nputative targets, adds an additional layer of complexity. From a clinical translational perspective, 234 \ngenetically-based predictions (e.g., SL interactions) are insufficient on their own to justify drug 235 \ndevelopment and rigorous validation through rigorous drug testing in model systems is essential to 236 \nbetter validate their therapeutic potential in humans. 237 \nWe then investigated whether cell lines with specific driver mutations demonstrate consistent 238 \nresponses to both genetic deletion and small-molecule inhibitors compared to their wild-type 239 \ncounterparts. We addressed this by analyzing drug sensitivities in cell lines stratified by driver mutation 240 \nstatus, examining both hotspot oncogene mutations and deleterious mutations in TSGs across 59 241 \ndriver genes (Table S1). Incorporating MSI as a confounder in the regression model was crucial for 242 \naccurately evaluating these associations, particularly for AKT inhibitors, which demonstrated high 243 \nefficacy in MSI-positive lines regardless of driver mutation status. Within our driver-focused SL network, 244 \n36 interaction pairs exhibited significant, directional agreement with drugs targeting the corresponding 245 \nprotein products (FDR > 0.2; Table 1). As expected, several of the strongest associations involved 246 \ntargeted oncology drugs in their anticipated genomic contexts—such as MDM2 inhibitors in TP53 wild-247 \ntype lines, CHK inhibitors in TP53-mutated lines [35], BRAF inhibitors in BRAF-  248 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n11 \n \nTable 1. Driver-target interaction pairs showing significant directional agreement between genetic dependencies 249 \nand drug sensitivities (FDR > 0.2). Each row lists the driver gene, target, interaction type (synthetic lethal or 250 \nresistant), genetic effect, and corresponding drug effect size and targets. Color gradients indicate the direction 251 \nand strength of genetic and drug effects. 252 \nmutated lines, and CDK4/6 inhibitors in RB1 wild-type lines. Furthermore, BRAF-, NRAS-, and KRAS-253 \nmutated lines exhibited marked resistance to PTPN11/SHP2 deletion and SHP2 inhibitor RMC-4550, 254 \nbut not other SHP2 inhibitors. PTEN-mutated lines showed heightened sensitivity to both PIK3CB 255 \ndeletion and several pan-PI3K inhibitors. One potential novel interaction emerged, as SIN3A mutated 256 \ncell lines were sensitive to nicotinamide phosphoribosyltransferase (NAMPT) deletion and associated 257 \nNAMPT inhibitor GMX1778, although functional links are unclear.  258 \nNext, we broadened our analysis to investigate all mutation-specific drug dependencies across the 259 \nPRISM database, aiming to uncover how often these sensitivities arise beyond our predefined genetic 260 \ninteraction pairs. Through this expanded analysis, we identified 44 significant interactions associated 261 \nwith tumor suppressor gene mutations and 96 with oncogenic mutations (Table S4). Our findings 262 \nconfirmed several known vulnerabilities, such as sensitivity to the Aurora kinase inhibitor LY3295668 in 263 \nSMARCA4-inactivated lines [36], and the SL of bafetinib – originally developed as a BCR-ABL and LYN 264 \ntyrosine kinase inhibitor – in BRAF-mutated cell lines. Notably, the latter association is supported by 265 \nrecent target deconvolution studies identifying BRAF as an off-target of bafetinib [37]. This validation 266 \nunderscores the utility of our approach in uncovering actionable vulnerabilities across diverse driver 267 \nmutations and drug combinations. Additionally, our analysis identified potentially novel drug 268 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n12 \n \nrepurposing opportunities, including the sensitivity of KDM6A-mutated cancer cell lines to the enoyl-acyl 269 \ncarrier protein reductase FABL inhibitor MUT056399 (Damaging; FDR = 0.00405) and the sensitivity of 270 \nFBXW7-mutated lines to PORCN inhibitor IWP-L6 (Hotspot; FDR = 0.025). FABL is a key enzyme in 271 \nthe bacterial fatty acid biosynthesis (FAS II) pathway that has garnered interest in antibacterial drug 272 \ndevelopment. However, metabolic dependencies, such as those disrupted by FABL inhibitors, are 273 \nincreasingly recognized in epigenetically mutated cancers, suggesting potential therapeutic relevance 274 \n[38].  275 \nCancer type-specific analysis 276 \nWe transitioned our efforts towards cancer type-specific interactions to identify context-dependent 277 \nopportunities for drug repurposing, as these may better reflect the unique genetic and tumor 278 \nmicroenvironment landscapes of different cancer types. Previous experimental knockout studies have 279 \nconsistently shown that SL gene pairs tend to exhibit cancer type specificity [39]. To explore these 280 \ninteractions, we repeated the Boruta genetic dependency analysis across 13 cancer type subsets 281 \n(Table S2), yielding an average of 4,528 interactions per cancer type. To prioritize the highest-value 282 \ninteractions, we applied stringent filtering criteria, including driver mutation status, mutation sample size 283 \n(n ≥ 4), and clustering based on importance scores using the heads/tails break algorithm, as previously 284 \ndescribed. This algorithm, specifically designed to address heavily tailed distributions, allowed us to 285 \nisolate the most significant interactions. After filtering, 2,500 interactions remained across 13 cancer 286 \ntypes, with 379 interactions including a druggable target (Fig S5). 287 \nApart from oncogene addiction-based SL interactions involving KRAS, and PIK3CA, and MDM2-TP53 288 \ninteractions, most high-confidence SL interactions were rarely shared across multiple cancer types (Fig 289 \nS6). This is likely due to context-specific dependencies, but it also reflects the challenges posed by 290 \ninsufficient sample sizes in some cancer types. Many cancer types lack the statistical power to identify 291 \nmutation stratified dependencies in cell lines screened for both KO and drug sensitivity assays. In many 292 \ncases, only a few highly mutated genes can be tested. Nevertheless, we identified some interactions 293 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n13 \n \nthat merit experimental follow-up. PBRM1-mutated renal cancer lines show resistance to BRD2 KO and 294 \nBET inhibitors. Similar findings have been made in triple-negative breast cancer models using CRISPR 295 \nKO and small molecule inhibitor screens [40].  296 \nWithin our bladder cancer SL network, the interaction between ARID1A and BRD2 emerged as an 297 \ninteresting candidate (Fig 6A). BRD2 has been previously highlighted as a therapeutic target in 298 \nARID1A mutated gastric cancer [41],  and ovarian clear cell carcinoma [42], suggesting it could also be 299 \nof clinical interest in bladder cancer as well. The ARID1A-BRD2 interaction was ranked on average 9.6 300 \nout of 47 significant SL interactions with matching drug targets in bladder cancer, combining ranks of 301 \nthe three importance algorithms (Table S2). Additionally, BRD2 knockout phenocopied BET inhibitor 302 \nsensitivity in bladder cancer cell lines as BRD2 essentiality and BET inhibitor sensitivity are strongly 303 \ncorrelated in both (R = 0.60; FDR = 8.15e-05; ). In agreement with the CRISPRi interaction, ARID1A 304 \nmutated bladder cancer lines were significantly more sensitive in 11 of 19 BET inhibitors – an effect that 305 \nwas not observed in other cancer types with recurrent ARID1A mutations (Wilcoxon; P < 0.05; Fig 6B). 306 \nA significant effect was also observed in the GDSC2 dataset, where bladder cancer cell lines showed 307 \nincreased sensitivity to OTX015 (P = 0.0574; Cohen’s d = -1.33). When observing the interaction in 308 \nmore detail, BRD2 essentiality was significantly enhanced in heterozygous mutated bladder cancer cell 309 \nlines (Fig 6C). In agreement, heterozygous ARID1A cell lines were also sensitive to several BET 310 \ninhibitors available in PRISM including ARV-825 and OTX015 (Fig 6D). Previous studies of mouse 311 \ngastric cancer in models have shown a similar dose dependent role of ARID1A where Arid1a-/+ tumors 312 \nfacilitate global loss of enhancers resulting in p53 suppression and tumor progression, whereas Arid1a-313 \n/- tumors initiates p53 activation and confers a competitive disadvantage [43].  314 \n3 Discussion 315 \nDespite extensive research efforts and the widespread adoption of high throughput KO and knockdown 316 \nscreens, PARP inhibitors (approved in 2014) remain the only available therapy leveraging SL. This 317 \nleaves a significant gap between discovery and clinical application. Approximately 90% of LoF driver 318 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n14 \n \nmutations currently lack SL-targeted therapies [44]. In this study, we systematically evaluated SL 319 \ninteractions involving cancer driver genes to assess their potential for drug repurposing opportunities 320 \nusing DepMap. Although many clinically focused, CRISPR-based, genetic SL interactions were highly 321 \nstatistically significant, only a subset translated effectively when mapped to drugs that target their gene 322 \nproduct, with most successful mappings corresponding to well-characterized interactions. These 323 \nfindings reveal some inherent limitations of pan-cancer SL analyses and emphasize the need for drug-324 \ngene assays to improve the translatability and clinical impact of SL-based drug development. 325 \nAssuming that CRISPR-based KO of a gene fully replicates the effects of pharmacological inhibition of 326 \nits encoded protein has motivated the use of CRISPR LoF screens as high-throughput proxies for 327 \nidentifying single-agent drugs with the potential to induce SL in cancer cells [45]. Although this 328 \napproach has achieved success in certain contexts [46], significant discrepancies between the 329 \noutcomes of genetic perturbation and pharmacological inhibition have been observed [47,48]. In our 330 \nanalysis, only 3% of drugs in the PRISM database showed a statistically significant correlation with 331 \nsurvival effects observed in the corresponding genetic KOs, with these correlations typically confined to 332 \nhighly targeted oncology agents. This discrepancy is not entirely unexpected. While small-molecule 333 \ninhibitors and CRISPR-mediated gene KOs can share functional similarities, they represent 334 \nfundamentally different modes of perturbation. Gene deletions result in the complete loss of gene 335 \nfunction and an abrupt cessation of expression. Although this may reveal acute cellular dependencies, 336 \nit also triggers compensatory responses such as the upregulation of paralogous genes [49], adaptive 337 \nrewiring of signaling pathways [50], and activation of the DNA damage response [51]. 338 \nIn contrast, small molecules typically target specific protein domains, often impairing catalytic activity 339 \nwhile leaving other functional aspects of the protein, such as transcriptional and translational regulation 340 \nor structural roles, intact. These drugs frequently target conserved domains shared across families of 341 \nstructurally related proteins. For example, BET inhibitors target bromodomains BD1 and BD2 present in 342 \nBRD2, BRD3, BRD4, and BRDT [52], and Imatinib is a competitive inhibitor of ATP binding to ABL 343 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n15 \n \nkinase and also targets the ATP-binding sites of other tyrosine kinases, including c-KIT and PDGFR 344 \n[53]. As a result, inhibitors often exhibit off-target effects that complicate the specificity of their action, 345 \nparticularly in kinase families, where entire groups of related proteins may be inadvertently inhibited. 346 \nThese off-target effects not only challenge therapeutic precision but can also confound efforts to identify 347 \nthe true biological drivers of anti-tumor responses [54]. In some cases, off-target effects serve as the 348 \nprimary mechanism of tumor suppression [55], highlighting the complexity of drug responses compared 349 \nto the complete LoF induced by gene KO’s. 350 \nWe believe that unbiased, high throughput methods such as CRISPR KO screens remain 351 \nindispensable for advancing SL drug development. However, integrating small molecules directly into 352 \ngenetic screens is likely crucial for capturing pharmacological effects and identifying clinically relevant 353 \nSL interactions, ultimately improving the translational relevance of these findings [34]. 354 \nAnother major challenge in translating SL interactions identified from in vitro screens such as DepMap, 355 \nas demonstrated by our work and that of others [1,11], is the pronounced context specificity of 356 \ndependencies. SL dependencies can vary widely depending on factors such as cell of origin, cancer 357 \ntype, cellular state, and experimental conditions, including media composition during screens [56]. This 358 \nvariability presents significant obstacles to identifying SL interactions with broad applicability across 359 \ndiverse cancer types, thereby limiting their translational potential. Controlling for these contextual 360 \nfactors often results in diminishing returns in the discovery of novel SL targets in pan-cancer analyses, 361 \nwhile cancer-specific analyses often suffer from insufficient statistical power. These conclusions reflect 362 \nthe inherent complexity of SL target discovery and highlights the need for: (i) accelerated molecular 363 \ncharacterization, genetic and drug sensitivity screening of cell lines across diverse cancer types and, (ii) 364 \ncontinued adoption of advanced screening strategies, such as integrating drug-gene 365 \n(pharmacogenomic), gene-gene (combinatorial CRISPR), and drug-drug (combinatorial drug) KO 366 \nscreens, to identify robust, context-specific vulnerabilities and therapeutic opportunities. 367 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n16 \n \nThe observed divergence between pharmacological inhibition and genetic knockout effects highlights 368 \nsignificant limitations of current large-scale genetic screens and the challenges of using genetic 369 \nperturbation to predict drug therapy effectiveness. These screens often fail to fully align drug sensitivity 370 \nprofiles with SL targets, emphasizing the need for a vastly expanded dependency database, alongside 371 \nmethodological advancements and integrative approaches to improve predictive accuracy. This aligns 372 \nwith the ambitious goal of comprehensively perturbing all protein-coding genes and evaluating the 373 \neffects of 10,000+ drugs across 20,000+ cancer models, as envisioned by recent initiatives [57]. 374 \nHowever, in their current state, these screening efforts are still constrained by the limited representation 375 \nof many common driver genes in specific cancer contexts and the lack of experimental models for rare 376 \ncancers. This highlights the need to expand and diversify experimental systems to build a more 377 \ncomprehensive dependency and drug sensitivity database to improve the predictive power of SL-based 378 \ndrug discovery. 379 \nOur study highlights key limitations in current SL target discovery approaches, including the 380 \npronounced context specificity of dependencies, the divergence between genetic KO’s and 381 \npharmacological inhibitors, and the limited representation of diverse cancer models. Addressing these 382 \nchallenges requires scaling up and diversifying experimental systems while leveraging advanced 383 \nstrategies such as isogenic screens, combinatorial CRISPR KO’s, and pharmacogenomic screens. The 384 \ntechnology to achieve this exists, but the ambition and resources to expand these efforts at scale are 385 \nnow critical to facilitate the translation of in vitro discoveries into clinical breakthroughs and accelerate 386 \nthe development of impactful precision cancer therapies. A recent pharmacogenomics study 387 \nexemplifies these challenges by conducting genetic KO screens targeting LoF drivers and evaluating 388 \ntheir interactions with clinically approved targeted therapies and cancer drugs. Of the 32 driver-drug 389 \ninteractions identified, only a KRAS hotspot mutations with RAF1 knockout was reproduced using 390 \nDepMap data (Pearson; P < 0.05; Fig. 7A). Interestingly, BRCA1 and STAG2 deletions sensitized cells 391 \nto PARP inhibitors in the pharmacogenomic screen; however, similar effects were absent in DepMap 392 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n17 \n \nwhen stratifying cell lines by BRCA1 and STAG2 mutation status. This discrepancy highlights the 393 \ncomplexity of PARP inhibitor mechanisms, which often rely on PARP trapping to stabilize PARP-DNA 394 \ncomplexes and induce cytotoxicity. Genetic KO’s, by contrast, do not replicate the formation of these 395 \ntoxic drug-target intermediates or their downstream effects on genomic instability and cell death. These 396 \nfindings emphasize the challenges of modeling pharmacogenetic mechanisms using genetic 397 \nperturbations alone and underscore the need to incorporate drug-specific effects into therapeutic 398 \ndevelopment. Expanding the scale and scope of experimental systems is essential to address these 399 \ngaps, improve the predictive power of SL-based drug discovery, and bridge the divide between in vitro 400 \nfindings and clinical applications.  401 \n 402 \n4 Methods 403 \nSynthetic lethal network construction 404 \nInput data 405 \nAll cell line data, including CRISPR-Cas9 knockout fitness scores, and mutation data, were obtained 406 \nfrom the DepMap portal (DepMap Public 23Q2; https://depmap.org/portal/). Mutation matrices were 407 \ndownloaded for both hotspot (OmicsSomaticMutationsMatrixHotspot.csv) and damaging mutations 408 \n(OmicsSomaticMutationsMatrixDamaging.csv). As described on the DepMap portal, DepMap 23Q2 409 \nmutations were called using Mutect2, with 0 indicating no mutation, 1 for heterozygous mutations, and 410 \n2 for homozygous mutations. Hotspot mutation sites were identified based on Hess et al. (2019). 411 \nCRISPR-Cas9 knockout effect scores were extracted from the CRISPRGeneEffect.csv file. For 412 \nDepMap 23Q2, gene-level knockout effect scores across all cell line models from both Achilles and 413 \nProject Score were computed using Chronos and harmonized using Harmonia. Microsatellite instability 414 \nscores were obtained from the DepMap Public 24Q2 release OmicsSignatures.csv file.  415 \nSelectively lethal genes  416 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n18 \n \nTo refine feature selection, we excluded genes that did not significantly affect cell proliferation across 417 \nall examined cell lines. Specifically, genes were retained in our dataset if they exhibited effect scores ≤ 418 \n-1 or ≥ 0.4 in at least four distinct cell lines. This approach ensures that only genes influencing cell 419 \nviability are included.  420 \nRandom forest feature selection. 421 \nTo limit the search space and reduce the burden of multiple test correction we implemented feature 422 \nselection classification using the Boruta algorithm (v8.0.0), and a workflow largely based on the PARIS 423 \npipeline (Benfatto et al., 2021). The Boruta algorithm works by first generating shadow features that are 424 \nrandom permutations of the original features. A random forest classifier is then employed to assess the 425 \nimportance of each original feature relative to the maximum importance of the permuted shadow 426 \nfeatures (shadowMax). If the importance of a feature – in this case gene mutation – significantly 427 \nexceeds shadowMax, it is retained; otherwise, it is rejected. In our analysis, Boruta was applied to each 428 \ngene using CRISPR-Cas9 knockout effect scores as the response variable, with concatenated hotspot 429 \nand damaging mutations as predictive features. Source genes with greater than 4 mutations were 430 \nincluded, as known SL interactions were observed as significant at this threshold (S7 Fig). The 431 \nalgorithm was run across all cell lines (pan-cancer) and on cancer-specific subsets 432 \n(OncotreePrimaryDisease) that had ≥ 28 fully characterized cell lines in CCLE. We employed three 433 \nimportance algorithms - getImpExtraGini, getImpRfZ, and getImpExtraRaw - to calculate feature 434 \nimportance within the random forest models (ranger v0.16.0). Interactions were considered significant 435 \nonly if all three algorithms concurred on feature importance. The significance threshold for feature 436 \nimportance was set at a p-value of 0.01, and the maximum number of iterations for Boruta was capped 437 \nat 500. All code used in this study will be publicly available on GitHub: 438 \nhttps://github.com/michaelcvermeulen/cancer_driver_repurposing (in progress) 439 \nDefining mutual exclusivity 440 \n 441 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n19 \n \nTumor mutation and copy number data was downloaded from TCGA.  ME was assessed for each 442 \ngenetic interaction pair using Fisher’s Exact Test, a statistical method to test the non-random 443 \nassociation between the mutation or deep deletion status of two genes across a cohort of tumor 444 \nsamples. For each pair, a 2x2 contingency table was constructed, with entries representing co-445 \noccurrence patterns: both genes mutated, only one gene mutated, or neither gene mutated. The 446 \nresulting p-values were used to determine significant ME, with the direction of association evaluated 447 \nusing odds ratios: OR < 1 indicated mutual exclusivity, while OR > 1 suggested co-occurrence. 448 \nTo account for cancer-type-specific effects, ME was tested separately in each TCGA cancer-type 449 \nsubset, with the most significant interaction retained for each interaction pair. Given the larger sample 450 \nsizes associated with driver mutations, we performed random sampling of non-driver gene pairs with 451 \nequal or larger sample sizes (n) to compare ME signal distributions. FDR correction using the 452 \nBenjamini-Hochberg method was applied to adjust for multiple hypothesis testing, with gene pairs 453 \nconsidered significant if FDR < 0.05. 454 \n 455 \nANOVA model 456 \nAll pan-cancer SL interaction pairs identified through Boruta analysis were further refined to assess 457 \ndirectionality, effect size, and potential confounding factors. Effect sizes were calculated using Cohen’s 458 \nd, while Pearson correlations were used to establish the directionality of interactions. To account for 459 \nconfounding variables, including microsatellite instability (MSI), cancer type, and cell line growth 460 \nproperties (adherent, suspension, or mixed), we applied a Type II ANOVA model adapted from Lord et 461 \nal. (2020). 462 \n 463 \nThe ANOVA model structure was defined as: 464 \n 465 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n20 \n \n𝐶𝑅𝐼𝑆𝑃𝑅 𝐸𝑓𝑓𝑒𝑐𝑡 𝑆𝑐𝑜𝑟𝑒 ~ 𝑀𝑆𝐼 + 𝐺𝑟𝑜𝑤𝑡ℎ 𝑃𝑟𝑜𝑝𝑒𝑟𝑡𝑖𝑒𝑠 + 𝐶𝑎𝑛𝑐𝑒𝑟 𝑇𝑦𝑝𝑒 + 𝑀𝑢𝑡𝑎𝑡𝑖𝑜𝑛 𝑆𝑡𝑎𝑡𝑢𝑠  471 \n 472 \nDamaging mutations were used for TSGs and hotspot mutations used for oncogenes. Mutation status 466 \np-values were extracted for each interaction and adjusted for multiple tests using Benjamini-Hochberg. 467 \nTo ensure the robustness of the final network, interactions were retained only if mutation status 468 \nexhibited an FDR < 0.05, excluding those where confounding factors contributed significantly to the 469 \nobserved effects. 470 \nIntersection with drug sensitivity data 473 \nPRISM drug sensitivity data were obtained from the DepMap portal (DepMap Public 23Q4 release). 474 \nGene targets for each drug in the PRISM dataset were compiled by merging information from Citeline 475 \nand PRISM repurposing target resources. For each pan-cancer synthetic lethal (SL) interaction, cell 476 \nlines were stratified based on the mutation status of the gene of interest. Drug sensitivity, specifically to 477 \nthose targeting the corresponding gene, was evaluated using Cohen’s effect size, Pearson correlation 478 \ncoefficients, and linear models, with adjustments for microsatellite instability (MSI) status, growth 479 \nproperties, and cancer type. Only cell lines from cancer types with more than 10 available lines in the 480 \nCCLE database, as classified by the OncotreePrimaryDisease field, were included in the analysis. 481 \nTo compare the effects of targeted drugs with the CRISPR-Cas9 knockout (KO) predictions, linear 482 \nmodels were fit between drug sensitivity profiles and CRISPR effect scores. These models were 483 \nadjusted for cancer type, MSI status, and growth properties to account for confounding factors when 484 \ntesting pan-cancer interactions.  485 \n 486 \nOverlap with pharmacogenomic data 487 \nNoteworthy pharmacogenomic interactions involving driver genes were sourced from Truesdell et al. 488 \n[34]. Each interaction involving targeted inhibitors was evaluated in both the DepMap and PRISM 489 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n21 \n \ndatasets by assessing target gene/protein proliferative effects stratified by driver gene mutation, using 490 \nboth (i) the pharmacological effect of the inhibitor and (ii) the impact of genetic knockout on the 491 \ncorresponding target. 492 \n 493 \nNetwork visualization  494 \nPan-cancer genetic interaction networks were visualized using the visNetwork package (v2.1.2) in R, 495 \noffering an interactive and customizable platform for exploring genetic interactions. Node sizes were 496 \nscaled based on total connections, while edge widths were proportional to the getImpExtraGini 497 \nimportance scores generated by Boruta. To enhance clarity, importance scores were capped at 30. 498 \nInteractions with druggable targets identified in PRISM were highlighted in blue. 499 \n 500 \nAcknowledgments 501 \nWe thank Christopher D Moyes for providing valuable guidance and feedback on the manuscript. We 502 \nthank Doris Coto Villa, Stephanie Young, Peter Truesdell and Ben Snider for valuable discussions and 503 \ncontributions to early stages of this work. 504 \nFunding 505 \nThis work was supported by a Canadian Institutes of Health Research Project Grant (PJT 178214) 506 \nawarded to Andrew Craig and Tomas Babak, and funds from the Ontario Graduate Scholarship 507 \nawarded to Michael Vermeulen. The funders had no role in study design, data collection and analysis, 508 \ndecision to publish, or preparation of the manuscript. 509 \n 510 \nAuthor Contributions 511 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n22 \n \nConceptualization – TB, AC 512 \nData curation – MV 513 \nFormal analysis – TB, MV 514 \nFunding acquisition – TB, AC, MV 515 \nMethodology – MV, TB 516 \nProject administration – TB, MV, AC  517 \nSupervision – TB, AC 518 \nVisualization - MV 519 \nWriting – original draft – MV, TB 520 \nWriting – review & editing – MV, TB, AC 521 \n 522 \nReferences 523 \n1.  O’Neil NJ, Bailey ML, Hieter P. Synthetic lethality and cancer. Nat Rev Genet. 2017;18: 613–623. 524 \ndoi:10.1038/nrg.2017.47 525 \n2.  Setton J, Zinda M, Riaz N, Durocher D, Zimmermann M, Koehler M, et al. Synthetic Lethality in Cancer 526 \nTherapeutics: The Next Generation. Cancer Discov. 2021;11: 1626–1635. doi:10.1158/2159-8290.CD-20-1503 527 \n3.  Rose M, Burgess JT, O’Byrne K, Richard DJ, Bolderson E. PARP Inhibitors: Clinical Relevance, Mechanisms 528 \nof Action and Tumor Resistance. Frontiers in Cell and Developmental Biology. 2020;8. Available: 529 \nhttps://www.frontiersin.org/articles/10.3389/fcell.2020.564601 530 \n4.  Arafeh R, Shibue T, Dempster JM, Hahn WC, Vazquez F. The present and future of the Cancer 531 \nDependency Map. Nat Rev Cancer. 2024. doi:10.1038/s41568-024-00763-x 532 \n5.  Srivatsa S, Montazeri H, Bianco G, Coto-Llerena M, Marinucci M, Ng CKY, et al. Discovery of synthetic 533 \nlethal interactions from large-scale pan-cancer perturbation screens. Nat Commun. 2022;13: 7748. 534 \ndoi:10.1038/s41467-022-35378-z 535 \n6.  Markowska M, Budzinska MA, Coenen-Stass A, Kang S, Kizling E, Kolmus K, et al. Synthetic lethality 536 \nprediction in DNA damage repair, chromatin remodeling and the cell cycle using multi-omics data from cell lines 537 \nand patients. Sci Rep. 2023;13: 7049. doi:10.1038/s41598-023-34161-4 538 \n7.  Karimpour M, Totonchi M, Behmanesh M, Montazeri H. Pathway-driven analysis of synthetic lethal 539 \ninteractions in cancer using perturbation screens. Life Sci Alliance. 2024;7: e202302268. 540 \ndoi:10.26508/lsa.202302268 541 \n8.  Rosenski J, Shifman S, Kaplan T. Predicting gene knockout effects from expression data. BMC Medical 542 \nGenomics. 2023;16: 26. doi:10.1186/s12920-023-01446-6 543 \n9.  Gonçalves E, Segura‐Cabrera A, Pacini C, Picco G, Behan FM, Jaaks P, et al. Drug mechanism‐of‐action 544 \ndiscovery through the integration of pharmacological and CRISPR screens. Mol Syst Biol. 2020;16: e9405. 545 \ndoi:10.15252/msb.20199405 546 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n23 \n \n10.  Lord CJ, Quinn N, Ryan CJ. Integrative analysis of large-scale loss-of-function screens identifies robust 547 \ncancer-associated genetic interactions. Murphy ME, editor. eLife. 2020;9: e58925. doi:10.7554/eLife.58925 548 \n11.  Tsherniak A, Vazquez F, Montgomery PG, Weir BA, Kryukov G, Cowley GS, et al. Defining a Cancer 549 \nDependency Map. Cell. 2017;170: 564-576.e16. doi:10.1016/j.cell.2017.06.010 550 \n12.  Corsello SM, Nagari RT, Spangler RD, Rossen J, Kocak M, Bryan JG, et al. Discovering the anticancer 551 \npotential of non-oncology drugs by systematic viability profiling. Nat Cancer. 2020;1: 235–248. 552 \ndoi:10.1038/s43018-019-0018-6 553 \n13.  Benfatto S, Serçin Ö, Dejure FR, Abdollahi A, Zenke FT, Mardin BR. Uncovering cancer vulnerabilities by 554 \nmachine learning prediction of synthetic lethality. Molecular Cancer. 2021;20: 111. doi:10.1186/s12943-021-555 \n01405-8 556 \n14.  Dempster JM, Krill-Burger JM, McFarland JM, Warren A, Boehm JS, Vazquez F, et al. Gene expression has 557 \nmore power for predicting in vitro cancer cell vulnerabilities than genomics. Cancer Biology; 2020 Feb. 558 \ndoi:10.1101/2020.02.21.959627 559 \n15.  Dou Y, Ren Y, Zhao X, Jin J, Xiong S, Luo L, et al. CSSLdb: Discovery of cancer-specific synthetic lethal 560 \ninteractions based on machine learning and statistic inference. Comput Biol Med. 2024;170: 108066. 561 \ndoi:10.1016/j.compbiomed.2024.108066 562 \n16.  Bazaga A, Leggate D, Weisser H. Genome-wide investigation of gene-cancer associations for the 563 \nprediction of novel therapeutic targets in oncology. Sci Rep. 2020;10: 10787. doi:10.1038/s41598-020-67846-1 564 \n17.  Shi X, Verduzco D, Petiwala S, Jeffries C, Lu C, Anton T, et al. TCGADEPMAP – Mapping Translational 565 \nDependencies and Synthetic Lethalities within The Cancer Genome Atlas. bioRxiv; 2022. p. 2022.03.24.485544. 566 \ndoi:10.1101/2022.03.24.485544 567 \n18.  He D, Liu Q, Wu Y, Xie L. A context-aware deconfounding autoencoder for robust prediction of 568 \npersonalized clinical drug response from cell-line compound screening. Nat Mach Intell. 2022;4: 879–892. 569 \ndoi:10.1038/s42256-022-00541-0 570 \n19.  Chiu Y-C, Zheng S, Wang L-J, Iskra BS, Rao MK, Houghton PJ, et al. Predicting and characterizing a cancer 571 \ndependency map of tumors with deep learning. Science Advances. 2021;7: eabh1275. 572 \ndoi:10.1126/sciadv.abh1275 573 \n20.  Cohen Z, Petrenko E, Barisaac AS, Abu-Zhayia ER, Yanovich-Ben-Uriel C, Ayoub N, et al. SLAYER: 574 \nSynthetic Lethality Analysis for Enhanced Targeted Therapy Implicates AhR inhibitor as a Target in RB1-Mutant 575 \nBladder Tumors. 2024. doi:10.1101/2024.05.01.592073 576 \n21.  Fedrizzi T, Ciani Y, Lorenzin F, Cantore T, Gasperini P, Demichelis F. Fast mutual exclusivity algorithm 577 \nnominates potential synthetic lethal gene pairs through brute force matrix product computations. 578 \nComputational and Structural Biotechnology Journal. 2021;19: 4394–4403. doi:10.1016/j.csbj.2021.08.001 579 \n22.  Liany H, Jayagopal A, Huang D, Lim JQ, Nbh NI, Jeyasekharan A, et al. ASTER: A Method to Predict 580 \nClinically Relevant Synthetic Lethal Genetic Interactions. IEEE J Biomed Health Inform. 2024;28: 1785–1796. 581 \ndoi:10.1109/JBHI.2024.3354776 582 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n24 \n \n23.  Srihari S, Singla J, Wong L, Ragan MA. Inferring synthetic lethal interactions from mutual exclusivity of 583 \ngenetic events in cancer. Biol Direct. 2015;10: 57. doi:10.1186/s13062-015-0086-1 584 \n24.  Yang W, Soares J, Greninger P, Edelman EJ, Lightfoot H, Forbes S, et al. Genomics of Drug Sensitivity in 585 \nCancer (GDSC): a resource for therapeutic biomarker discovery in cancer cells. Nucleic Acids Research. 2013;41: 586 \nD955–D961. doi:10.1093/nar/gks1111 587 \n25.  Carli F, Chiaro PD, Morelli M, Arora C, Bisceglia L, Rosa NDO, et al. Learning and actioning general 588 \nprinciples of cancer cell drug sensitivity. bioRxiv; 2024. p. 2024.03.28.586783. doi:10.1101/2024.03.28.586783 589 \n26.  Baek B, Jang E, Park S, Park S-H, Williams DR, Jung D-W, et al. Integrated drug response prediction 590 \nmodels pinpoint repurposed drugs with effectiveness against rhabdomyosarcoma. PLOS ONE. 2024;19: 591 \ne0295629. doi:10.1371/journal.pone.0295629 592 \n27.  Pellecchia S, Viscido G, Franchini M, Gambardella G. Predicting drug response from single-cell expression 593 \nprofiles of tumours. BMC Medicine. 2023;21: 476. doi:10.1186/s12916-023-03182-1 594 \n28.  Goodspeed A, Heiser LM, Gray JW, Costello JC. Tumor-Derived Cell Lines as Molecular Models of Cancer 595 \nPharmacogenomics. Molecular Cancer Research. 2016;14: 3–13. doi:10.1158/1541-7786.MCR-15-0189 596 \n29.  Li A, Walling J, Kotliarov Y, Center A, Steed ME, Ahn SJ, et al. Genomic Changes and Gene Expression 597 \nProfiles Reveal That Established Glioma Cell Lines Are Poorly Representative of Primary Human Gliomas. 598 \nMolecular Cancer Research. 2008;6: 21–30. doi:10.1158/1541-7786.MCR-07-0280 599 \n30.  Salvadores M, Fuster-Tormo F, Supek F. Matching cell lines with cancer type and subtype of origin via 600 \nmutational, epigenomic, and transcriptomic patterns. Science Advances. 2020;6: eaba1862. 601 \ndoi:10.1126/sciadv.aba1862 602 \n31.  Bailey MH, Tokheim C, Porta-Pardo E, Sengupta S, Bertrand D, Weerasinghe A, et al. Comprehensive 603 \nCharacterization of Cancer Driver Genes and Mutations. Cell. 2018;173: 371-385.e18. 604 \ndoi:10.1016/j.cell.2018.02.060 605 \n32.  Campbell PJ, Getz G, Korbel JO, Stuart JM, Jennings JL, Stein LD, et al. Pan-cancer analysis of whole 606 \ngenomes. Nature. 2020;578: 82–93. doi:10.1038/s41586-020-1969-6 607 \n33.  Nulsen J, Misetic H, Yau C, Ciccarelli FD. Pan-cancer detection of driver genes at the single-patient 608 \nresolution. Genome Med. 2021;13: 12. doi:10.1186/s13073-021-00830-0 609 \n34.  Truesdell P, Chang J, Coto Villa D, Dai M, Zhao Y, McIlwain R, et al. Pharmacogenomic discovery of 610 \ngenetically targeted cancer therapies optimized against clinical outcomes. npj Precis Onc. 2024;8: 1–13. 611 \ndoi:10.1038/s41698-024-00673-z 612 \n35.  Zemanova J, Hylse O, Collakova J, Vesely P, Oltova A, Borsky M, et al. Chk1 inhibition significantly 613 \npotentiates activity of nucleoside analogs in TP53-mutated B-lymphoid cells. Oncotarget. 2016;7: 62091–62106. 614 \ndoi:10.18632/oncotarget.11388 615 \n36.  Tagal V, Wei S, Zhang W, Brekken RA, Posner BA, Peyton M, et al. SMARCA4-inactivating mutations 616 \nincrease sensitivity to Aurora kinase A inhibitor VX-680 in non-small cell lung cancers. Nat Commun. 2017;8: 617 \n14098. doi:10.1038/ncomms14098 618 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n25 \n \n37.  Ji H, Lu X, Zhao S, Wang Q, Liao B, Bauer LG, et al. Target deconvolution with matrix-augmented pooling 619 \nstrategy reveals cell-specific drug-protein interactions. Cell Chem Biol. 2023;30: 1478-1487.e7. 620 \ndoi:10.1016/j.chembiol.2023.08.002 621 \n38.  Xu X, Peng Q, Jiang X, Tan S, Yang Y, Yang W, et al. Metabolic reprogramming and epigenetic 622 \nmodifications in cancer: from the impacts and mechanisms to the treatment potential. Exp Mol Med. 2023;55: 623 \n1357–1370. doi:10.1038/s12276-023-01020-1 624 \n39.  Wooller SK, Pearl LH, Pearl FMG. Identifying actionable synthetically lethal cancer gene pairs using 625 \nmutual exclusivity. FEBS Lett. 2024;598: 2028–2039. doi:10.1002/1873-3468.14950 626 \n40.  Shu S, Wu H-J, Ge JY, Zeid R, Harris IS, Jovanović B, et al. Synthetic Lethal and Resistance Interactions 627 \nwith BET Bromodomain Inhibitors in Triple-Negative Breast Cancer. Mol Cell. 2020;78: 1096-1113.e8. 628 \ndoi:10.1016/j.molcel.2020.04.027 629 \n41.  Geng H, Qian R, Zhong Y, Tang X, Zhang X, Zhang L, et al. Leveraging synthetic lethality to uncover 630 \npotential therapeutic target in gastric cancer. Cancer Gene Ther. 2024;31: 334–348. doi:10.1038/s41417-023-631 \n00706-y 632 \n42.  Berns K, Caumanns JJ, Hijmans EM, Gennissen AMC, Severson TM, Evers B, et al. ARID1A mutation 633 \nsensitizes most ovarian clear cell carcinomas to BET inhibitors. Oncogene. 2018;37: 4611–4625. 634 \ndoi:10.1038/s41388-018-0300-6 635 \n43.  Loe AKH, Francis R, Seo J, Du L, Wang Y, Kim J-E, et al. Uncovering the dosage-dependent roles of Arid1a 636 \nin gastric tumorigenesis for combinatorial drug therapy. Journal of Experimental Medicine. 2021;218: 637 \ne20200219. doi:10.1084/jem.20200219 638 \n44.  Hay M, Thomas DW, Craighead JL, Economides C, Rosenthal J. Clinical development success rates for 639 \ninvestigational drugs. Nat Biotechnol. 2014;32: 40–51. doi:10.1038/nbt.2786 640 \n45.  Fellmann C, Gowen BG, Lin P-C, Doudna JA, Corn JE. Cornerstones of CRISPR–Cas in drug discovery and 641 \ntherapy. Nat Rev Drug Discov. 2017;16: 89–100. doi:10.1038/nrd.2016.238 642 \n46.  Xia Y, Ji X, Jang IS, Surka C, Hsu C, Wang K, et al. Genetic and pharmacological interrogation of cancer 643 \nvulnerability using a multiplexed cell line screening platform. Commun Biol. 2021;4: 834. doi:10.1038/s42003-644 \n021-02352-2 645 \n47.  DeWeirdt PC, Sangree AK, Hanna RE, Sanson KR, Hegde M, Strand C, et al. Genetic screens in isogenic 646 \nmammalian cell lines without single cell cloning. Nat Commun. 2020;11: 752. doi:10.1038/s41467-020-14620-6 647 \n48.  Lin A, Giuliano CJ, Sayles NM, Sheltzer JM. CRISPR/Cas9 mutagenesis invalidates a putative cancer 648 \ndependency targeted in on-going clinical trials. Settleman J, editor. eLife. 2017;6: e24179. 649 \ndoi:10.7554/eLife.24179 650 \n49.  Rossi A, Kontarakis Z, Gerri C, Nolte H, Hölper S, Krüger M, et al. Genetic compensation induced by 651 \ndeleterious mutations but not gene knockdowns. Nature. 2015;524: 230–233. doi:10.1038/nature14580 652 \n50.  El-Brolosy MA, Kontarakis Z, Rossi A, Kuenne C, Günther S, Fukuda N, et al. Genetic compensation 653 \ntriggered by mutant mRNA degradation. Nature. 2019;568: 193–197. doi:10.1038/s41586-019-1064-z 654 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n26 \n \n51.  Haapaniemi E, Botla S, Persson J, Schmierer B, Taipale J. CRISPR–Cas9 genome editing induces a p53-655 \nmediated DNA damage response. Nat Med. 2018;24: 927–930. doi:10.1038/s41591-018-0049-z 656 \n52.  Filippakopoulos P, Qi J, Picaud S, Shen Y, Smith WB, Fedorov O, et al. Selective inhibition of BET 657 \nbromodomains. Nature. 2010;468: 1067–1073. doi:10.1038/nature09504 658 \n53.  Druker BJ, Talpaz M, Resta DJ, Peng B, Buchdunger E, Ford JM, et al. Efficacy and Safety of a Specific 659 \nInhibitor of the BCR-ABL Tyrosine Kinase in Chronic Myeloid Leukemia. New England Journal of Medicine. 660 \n2001;344: 1031–1037. doi:10.1056/NEJM200104053441401 661 \n54.  Green MR, Newton MD, Fancher KM. Off-Target Effects of BCR-ABL and JAK2 Inhibitors. Am J Clin Oncol. 662 \n2016;39: 76–84. doi:10.1097/COC.0000000000000023 663 \n55.  Lin A, Giuliano CJ, Palladino A, John KM, Abramowicz C, Yuan ML, et al. Off-target toxicity is a common 664 \nmechanism of action of cancer drugs undergoing clinical trials. Sci Transl Med. 2019;11: eaaw8412. 665 \ndoi:10.1126/scitranslmed.aaw8412 666 \n56.  Hart T, Chandrashekhar M, Aregger M, Steinhart Z, Brown KR, MacLeod G, et al. High-Resolution CRISPR 667 \nScreens Reveal Fitness Genes and Genotype-Specific Cancer Liabilities. Cell. 2015;163: 1515–1526. 668 \ndoi:10.1016/j.cell.2015.11.015 669 \n57.  Boehm JS, Garnett MJ, Adams DJ, Francies HE, Golub TR, Hahn WC, et al. Cancer research needs a better 670 \nmap. Nature. 2021;589: 514–516. doi:10.1038/d41586-021-00182-0 671 \n 672 \n 673 \nSupporting information captions 674 \n 675 \nS1 Table. DepMap Pan-cancer SL network interactions and related PRISM drug correlations 676 \nS2 Table. DepMap Cancer-type specific SL network interactions and related PRISM drug correlations  677 \n 678 \nS3 Table. Correlations between CRISPR effect score and PRISM drug sensitivity 679 \nS4 Table. Driver gene-stratified differential PRISM drug sensitivity for damaging and hotspot mutations 680 \nS1 Fig. Schematic illustrating genetic interaction concepts relevant to cancer genomics. 681 \nS2 Fig. Distribution of synthetic lethal (SL) interaction significance values stratified by mutation 682 \nfrequency across cell lines. 683 \nS3 Fig. Overview of pan-cancer SL and resistance genetic interactions identified in the pan-cancer 684 \nanalysis. 685 \nS4 Fig. Correlation of drug sensitivity with genetic dependency profiles for MCL1 inhibitors  686 \nS5 Fig. Overview of SL interactions identified across cancer types following filtering criteria. 687 \nS6 Fig. Top cancer-specific SL interactions identified across CCLE cell lines. 688 \nS7 Fig. Relationship between the statistical significance of SL genetic interactions and the number of 689 \nmutated cell lines. 690 \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \nFigure 1. An overview of the approach to select for clinically actionable targeted drug repurposing \nopportunities. Using the Boruta algorithm, our pipeline integrates data from the DepMap consortium (release \n23Q2) to construct a genetic interaction network. Fitness scores from CRISPR loss-of-function screens serve \nas dependent variables, while cell line mutation data (hotspot and damaging) are used as independent \nvariables in a random forest (RF) feature selection process. For each perturbed gene, the algorithm assigns an \nimportance score to mutation features, retaining those with any predictive value, and forming the foundation of \nour synthetic lethality (SL) network. To establish data features that are predictive of interactions holding up in \nthe clinic (i.e. tumors), we utilized TCGA mutation data to assess mutual exclusivity (ME) across sufficiently \npowered genes.Non-overlapping mutations in human tumors suggest incompatibility and potential synthetic \nlethality. The resulting genetic interaction network was then integrated with PRISM drug screen data (release \n23Q4) to identify drugs that mirror CRISPR effect scores and display differential sensitivity in driver-stratified \nsubsets, generating a list of candidate drugs for targeted repurposing. \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n  \nFigure 2. Cancer driver based genetic interaction network.  \nA) Force-directed network of 1,428 high-confidence genetic interaction pairs identified from DepMap CRISPR \nviability screens, accounting for biases arising from cancer type and microsatellite instability. The network \nincludes both synthetic lethal and synthetic viability interactions. B) Genetic interaction network filtered to show \ndriver source mutations, with driver gene nodes colored red (oncogenes) and blue (tumor suppressor genes) \nand non-drivers in yellow. Blue edges indicate genetic interactions containing a drug target listed in the PRISM \ndatabase, with edge width representing the strength of the interaction (Importance value). Node size reflects \nthe total number of connections. C) Comparison of mutual exclusivity significance (y-axis) in TCGA data \nbetween synthetic lethal pairs containing a driver gene (blue) versus non-driver interactions (red) shows that \ninteractions involving driver genes are more likely to be mutually exclusive. Grey lines represent the distribution \nof randomly permuted subsets of non-driver interactions. \n \n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n \n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \nFigure 3.  \nA) Heatmap depicting effect sizes of pan-cancer SL interactions involving driver genes, derived from DepMap \nCRISPR effect scores (ANOVA; FDR < 0.1). Driver mutations are shown on the x-axis, and CRISPR target \ngenes on the y-axis. Negative effect size values indicate that mutated drivers sensitize cells to target gene \nknockout, while positive values suggest that driver mutations confer resistance or that wild-type alleles \nsensitize cells. On the right, Pearson correlations between PRISM drug sensitivity and CRISPR knockout \neffects for sample interactions are provided. Higher Pearson correlation significance indicates alignment \nbetween drug sensitivity and CRISPR knockout profiles across all CCLE cell lines, implying comparable \nimpacts on cell viability from drug inhibition and gene knockout. Drugs with negative correlations are labeled in \nblue. B) Selected examples are highlighted in white-outlined boxes: Box 1, BRAF; Box 4, PIK3CA; Box 5, \nCDK6/RB1. C) Known SL interactions, including paralog pairs, are provided as positive controls: Box 3, \nSMARCA2/4; Box 2, STAG1/2. \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \nFigure 4: Drug perturbations typically do not replicate genetic perturbations of corresponding drug \ntargets. \nDensity plot showing the distribution of Spearman's rank correlation coefficients between gene effect scores \nand PRISM drug sensitivity for drugs with overlapping gene targets (blue). The red curve represents the \nSpearman's rank correlation coefficients between each PRISM drug and a set of randomly permuted gene \neffect scores, serving as a control. Bottom: Boxplot summarizing the same data, highlighting some of the \nstrongest correlations between gene effects and drug sensitivity.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n \n \nFigure 5: Pan-cancer synthetic lethal interactions recapitulated by PRISM drug sensitivity.  \nSynthetic lethal interactions involving hotspot and damaging driver mutations were overlapped with PRISM \ndrugs targeting the same genes. Tumor suppressor genes and oncogenes are indicated in blue and red, \nrespectively. The inset provides a magnified view of the central plot region. For each SL interaction, cell lines \nwere categorized as wild-type or mutant for the source gene, and differences between CRISPR effect scores \n(y-axis) and PRISM drug logFC values (x-axis) were assessed using type II ANOVA, controlling for cancer type \nand MSI status. Text labels follow the format \"drug_source-mutation_target gene.\" P-values were -log10 \ntransformed and adjusted by the sign of the effect size to reflect directionality. Points in the top right quadrant \nhave a positive effect size for both drug sensitivity and CRISPR effect, indicating decreased viability in wild-\ntype lines for the driver gene, while points in the bottom left quadrant show increased sensitivity in driver-\nmutant lines to both drug and CRISPR knockout. For instance, REFAMETINIB_BRAF_MAP2K1 in the bottom \nleft quadrant demonstrates that BRAF-mutant cell lines are sensitive to both MAP2K1 CRISPR deletion and \nthe MAP2K1 inhibitor Refametinib. Points in the top left and bottom right suggest discordance between drug \nand CRISPR effects on viability. When multiple PRISM drugs matched a target gene, the most significant \ninteraction was plotted. Full data are available in S4 Table.  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n \n \n \n \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \nFigure 6. Overlap of DepMap synthetic lethal (SL) interactions and PRISM drug sensitivity in bladder \nurothelial carcinoma, highlighting ARID1A/BRD2 as a potential repurposing opportunity. \nA) SL interactions identified by CRISPR screens and corresponding drug sensitivities from the PRISM \ndatabase in bladder urothelial carcinoma cell lines. The x-axis indicates the significance of genetic \ndependencies (CRISPR-Mut), and the y-axis shows the significance of drug sensitivity differences (Drug-Mut), \nrepresented as -log10(p) values. Circle size denotes sample size, while labeled red points highlight significant \ndrug-gene interactions, notably the ARID1A-BRD2 interaction with BET inhibitor ARV-825. The top gene-drug \ninteractions are shown when multiple drugs have the same predicted targets. B) Heatmap depicting differential \nsensitivity of bladder cancer cell lines carrying ARID1A mutations versus wild-type cell lines to various BET \ninhibitors from the PRISM dataset. ARID1A-mutated cell lines demonstrate consistent and heightened \nsensitivity to nearly all BET inhibitors in bladder cancer. C) Boxplot showing significantly lower BRD2 knockout \n(KO) fitness scores in bladder cancer cell lines harboring damaging ARID1A mutations compared to wild-type \nlines, suggesting increased dependency on BRD2 in the heterozygous mutant context. D) Boxplot illustrating \nheightened sensitivity (negative fold change in viability scores from PRISM primary screening) of ARID1A-\nmutated bladder cancer cell lines to the BET inhibitor ARV-825 compared to wild-type cell lines. \n  \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint \n\n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \n \nFigure 7. Limited replication of driver-drug interactions between pharmacogenomic and large-scale \ngenomic screens. (A) Pearson correlation coefficients for the association between driver mutations and \nknockout of drug targets in DepMap CRISPR screens, stratified by interaction type (sensitizing vs. \nsuppressive) that were detected in pharmacogenomic screens (Truesdell et al., 2024). (B) Corresponding \ncorrelations for drug sensitivity profiles from PRISM screens. Although KRAS–RAF1 was reproduced in \nDepMap (A), most interactions, including BRCA1–PARP1 and STAG2–RUCAPARIB, did not show consistent \npatterns across datasets. \n \n \n \n.CC-BY 4.0 International licenseavailable under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made \nThe copyright holder for this preprintthis version posted March 22, 2025. ; https://doi.org/10.1101/2025.03.20.644275doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}