RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: results of the EDIT-B consortium study

RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: results of the EDIT-B consortium study | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (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];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: results of the EDIT-B consortium study View ORCID Profile D. Weissmann , View ORCID Profile N. Salvetat , C. Cayzac , View ORCID Profile F. Checa Robles , View ORCID Profile D. Vetter , L. Walczer-Baldinazzo , G. Ruggeri , M. Ferrari , View ORCID Profile A. Miranda-Mendizabal , J. Zambrano , J. Zarp , View ORCID Profile A. Giménez-Palomo , View ORCID Profile M. Valenti , View ORCID Profile JM. Haro , View ORCID Profile LV. Kessing , View ORCID Profile C. Henry , View ORCID Profile E. Vieta doi: https://doi.org/10.1101/2025.11.05.25339583 D. Weissmann a ALCEDIAG / Sys2Diag, CNRS UMR 9005, Parc Euromédecine , Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for D. Weissmann N. Salvetat a ALCEDIAG / Sys2Diag, CNRS UMR 9005, Parc Euromédecine , Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for N. Salvetat For correspondence: nicolas.salvetat{at}alcediag-alcen.com C. Cayzac a ALCEDIAG / Sys2Diag, CNRS UMR 9005, Parc Euromédecine , Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site F. Checa Robles a ALCEDIAG / Sys2Diag, CNRS UMR 9005, Parc Euromédecine , Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for F. Checa Robles D. Vetter a ALCEDIAG / Sys2Diag, CNRS UMR 9005, Parc Euromédecine , Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for D. Vetter L. Walczer-Baldinazzo b SYNLAB Italia – Genetic Unit , via Beato L. Pavoni 18, 25014, Castenedolo (BS), Italy Find this author on Google Scholar Find this author on PubMed Search for this author on this site G. Ruggeri b SYNLAB Italia – Genetic Unit , via Beato L. Pavoni 18, 25014, Castenedolo (BS), Italy Find this author on Google Scholar Find this author on PubMed Search for this author on this site M. Ferrari c IRCCS, SYNLAB SDN , Via Gianturco 113, Naples, Italy Find this author on Google Scholar Find this author on PubMed Search for this author on this site A. Miranda-Mendizabal d Parc Sanitari Sant Joan de Déu, Sant Joan de Déu Research Institute , Dr. Antoni Pujadas, 42 Sant Boi de Llobregat (Barcelona), Spain i Mental Health Networking Biomedical Research Centre (CIBERSAM) , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for A. Miranda-Mendizabal J. Zambrano g Université de Paris Cité, Paris, France; GHU Psychiatrie & Neurosciences , Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site J. Zarp e Copenhagen Affective Disorder Research Centre (CADIC), Psychiatric Centre Copenhagen , Copenhagen, Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site A. Giménez-Palomo h Bipolar and Depressive Disorders Unit, Hospital Clinic , Barcelona, Spain i Mental Health Networking Biomedical Research Centre (CIBERSAM) , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for A. Giménez-Palomo M. Valenti h Bipolar and Depressive Disorders Unit, Hospital Clinic , Barcelona, Spain i Mental Health Networking Biomedical Research Centre (CIBERSAM) , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for M. Valenti JM. Haro d Parc Sanitari Sant Joan de Déu, Sant Joan de Déu Research Institute , Dr. Antoni Pujadas, 42 Sant Boi de Llobregat (Barcelona), Spain i Mental Health Networking Biomedical Research Centre (CIBERSAM) , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for JM. Haro LV. Kessing e Copenhagen Affective Disorder Research Centre (CADIC), Psychiatric Centre Copenhagen , Copenhagen, Denmark f Department of Clinical Medicine, University of Copenhagen , Denmark Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for LV. Kessing C. Henry g Université de Paris Cité, Paris, France; GHU Psychiatrie & Neurosciences , Paris, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for C. Henry E. Vieta h Bipolar and Depressive Disorders Unit, Hospital Clinic , Barcelona, Spain i Mental Health Networking Biomedical Research Centre (CIBERSAM) , Madrid, Spain Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for E. Vieta Abstract Full Text Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Background Differentiating bipolar disorder (BD) from major depressive disorder (MDD) during depressive episodes represents one of the most critical challenges in clinical psychiatry as patients may not recall prior hypomanic or manic episodes. RNA editing biomarkers combined with AI may improve diagnostic accuracy and support personalized therapeutic approaches. Methods The assay used is an in vitro diagnostic (IVD) blood test based on 8 genes that leverages adenosine-to-inosine (A-to-I) RNA editing signatures to aid in the differentiation of BD from MDD. A set of RNA-editing blood biomarkers within these 8 genes was previously identified and published. In this multicenter European study, we assessed the performance of three distinct RNA editing–based algorithms developed prior to this project. Sequencing data were processed via a secure HDS–certified platform by central clinical laboratory and the results for each algorithm were compared against physician’s diagnosis to evaluate diagnostic performance. Results Across 426 patients with a current major depressive episode recruited from four European centers, the three algorithms demonstrated robust diagnostic performance. Algorithm A284, exhibited the strongest results and highest performance, with high sensitivity (82.6%) and specificity (80.3%) across sites and patient subgroups, thus reproducing the findings and performance obtained on a previous independent cohort. Conclusion This study validates A-to-I RNA editing signatures coupled with AI as a reliable approach to distinguish BD from MDD highlighting the clinical utility of the assay and algorithm A284 as an innovative, robust, complementary diagnostic tool. 1. Introduction The differential diagnosis of bipolar disorder (BD) and major depressive disorder (MDD) remains a critical unmet need in psychiatry. Although both conditions present with depressive episodes, their pathophysiological underpinnings and treatment trajectories differ substantially. The initial presentation of BD is often depressive, leading to frequent misdiagnosis as MDD and an average diagnostic delay of nearly a decade( 1 – 5 ). This delay carries significant clinical consequences, including inappropriate pharmacotherapy, increased risk of manic switching, and elevated suicide rates( 6 – 11 ). Traditional diagnostic approaches, relying on symptomatology and structured interviews, lack the precision required to disentangle overlapping clinical phenotypes( 12 ). While certain features—early onset, mood lability, familial risk—may signal BD, their predictive value is limited. The absence of objective, biologically grounded tools for differential diagnosis continues to impede early and accurate identification. Recent efforts in molecular psychiatry have explored immune and oxidative pathways( 13 – 15 ), yet robust, reproducible biomarkers remain elusive( 16 , 17 ). Among emerging candidates, RNA editing—a dynamic epitranscriptomic mechanism altering nucleotide sequences post-transcription—has shown promise. Dysregulated editing patterns have been reported in BD and other neuropsychiatric disorders, implicating this process in mood regulation and neuroinflammation( 18 – 31 ). A recent study identified a blood-based RNA editing signature, including a panel of 8 RNA sequences and an algorithm, capable of distinguishing BD from MDD with high sensitivity and specificity( 25 ). This molecular signature, integrated with demographic features via AI-based classification, demonstrated strong diagnostic performance in both discovery and independent validation cohorts. The EDIT-B project is funded by the European Union (Grant 220628/230125). The primary objective of the EDIT-B project is to externally validate and compare the diagnostic performance of three previously identified algorithms, which operate on the above-described panel of 8 RNA sequences combined with the clinical data as input data for differentiating BD from MDD in individuals currently experiencing major depressive episodes( 32 ). The specific aim of this study is to determine which of the three RNA editing signatures demonstrates the highest diagnostic accuracy (sensitivity, specificity, and AUC-ROC) in an independent clinical sample. 2. MATERIALS AND METHODS The protocol of this study has been published in Miranda et al, Annals of General Psychiatry, ( 2025 ) , 24:7, https://doi.org/10.1186/s12991lJ024lJ00544lJ8 . 2.1. Study design and setting The present trial is a multicenter cross-sectional study enrolling adults diagnosed with BD or MDD with a current major depressive episode across four clinical sites: two in Barcelona (Spain), one in Paris (France), and one in Copenhagen (Denmark) (Suppl Table 1). The inclusion of geographically and demographically diverse European centers was intended to enhance the representativeness and generalizability of findings. The study is registered on ClinicalTrials.gov (identifier: NCT05603819 ) ( Table 1 ). Reporting followed the SPIRIT guidelines, with the corresponding checklist provided( 33 ) (Suppl Table 1). The final analysis was performed based on the comparison of clinicians’ diagnosis informed on the electronic case report form (eCRF) vs IVD blood test results at ALCEDIAG after the EDIT-B project database freezing. View this table: View inline View popup Download powerpoint Table 1: Demographic and clinical characteristics of the participant included in the Analytical Set. Data are the mean ± SEM. p-values of main characteristics are obtained with the Student’s t-test or chi-squared test. MDD: Major depressive disorder patients; BD: bipolar disorder patients; MADRS: Montgomery-Åsberg depression rating scale; YMRS: Young Mania Rating Scale; BMI: Body Mass Index. A central clinical laboratory (SYNLAB Italia – Genetic Unit, via Beato L. Pavoni 18, 25014, Castenedolo (BS), Italy) processed the blood samples (Suppl Table 1). All the biological processing of the blood samples and the use of algorithms allowing the determination of the categorization were made blindly, i.e without any knowledge of the clinicians’ diagnoses. In addition, both global and individual results of the signatures were not communicated to the investigators before the end of the study. This study is conducted in accordance with the ethical principles set out in the latest version of the World Medical Association’s Declaration of Helsinki 2013, attending to all the nuances involved in precision psychiatry research and following the requirements of good clinical practice( 34 ). Patients were not involved when designing the study, but the study investigators actively participated in the study design and setting of the research question. 2.2 Sample size calculation The sample size was initially fixed to provide a sufficient precision in the estimates of the test performances in terms of sensitivity, specificity, accuracy, and AUC-ROC. A total sample size of N=392 patients with depression with a 1:1: ratio of MDD (unipolar) vs. BD was calculated to allow the estimation of the four performance parameters with a precision of ± 5% or better (considering that the value of specificity is the lowest of the four expected values of performances parameters, precision of the estimates was better for these three latter). We used the methods proposed by Buderer and Jones for these calculations( 35 , 36 ). Considering a maximal 10% attrition rate, 436 patients were supposed to be included. The working hypothesis of this study is the reproducibility of the results previously achieved, i.e. 80.6%, 85.1%, 83.7%, and 0.901 respectively. 2.3 Clinical Assessment and Eligibility criteria Inclusion criteria Eligible participants were adults aged 18–80 years with a diagnosis of BD or MDD, currently experiencing a MDE. Diagnoses were confirmed according to DSM-5 criteria using the Mini-International Neuropsychiatric Interview (MINI)( 37 ). Both inpatients and outpatients were eligible. A Montgomery–Åsberg Depression Rating Scale (MADRS) score( 38 ) equal to or greater than 20 at assessment was required to be included. BD participants must have had at least one prior manic or hypomanic episode, while MDD participants must have had at least one previous MDE. Written informed consent was obtained from all participants after oral and written explanation of the study Exclusion criteria The exclusion criteria for MDD participants are having first-degree family history of BD (e.g., parents, siblings or children), a total score on the Young Mania Rating Scale (YMRS)( 39 ) equal to or greater than 12 at the time of assessment, pregnancy, BD or MDD secondary to major central nervous system affect or a diagnosis of schizoaffective disorder ( Fig. 1 ). Download figure Open in new tab Figure 1. STARD flow diagram of participants included in the EDIT-B consortium study. The EDIT-B consortium study is a multicenter study recruiting participants in four different centers in Spain, France and Denmark, respectively. Participants are assessed for eligibility with several inclusion and exclusion criteria during the inclusion visit. Download figure Open in new tab Figure 2: Summary of the participant flow in the EDIT-B consortium study. (1) Participants were enrolled by the clinician and received a diagnostic evaluation; (2) After inclusion, a blood sample was collected at the respective study centre; (3) Samples were shipped to the central laboratory for performs the RNA editing-based IVD test; (4) Data were analyzed using the three algorithms via the EDIT-B project platform (5) Clinical eCRF data and IVD test results were compared in the final statistical analysis of the analytical set. Cohort description Recruitment took place between summer 2022 and September 2024. A total of 426 participants were screened for eligibility: 259 with MDD (60.8%), 165 with BD (38.7%), and 2 (0.5%) with missing diagnostic data ( Figure 1 ). Eight participants (1.9%) did not meet the screening inclusion/exclusion criteria: 4 with MDD, 2 with BD, and 2 with indeterminate diagnosis. From the 418 individuals meeting inclusion criteria (inclusion set), 14 were excluded for protocol deviations (3 missing clinician signatures on consent forms, 1 with family psychiatric history, 1 with CNS disease, and 9 with improperly handled blood samples). During the assay processing, 14 additional participants were excluded: 4 without psychiatric treatment (not compliant with the intended use), 2 lacking sequencing data, and 8 not meeting technical quality control thresholds (sequencing depth <10 000 reads). Finally, the analytical set included 393 participants: 238 with MDD (60.6%) and 155 with BD (39.4%). 2.4 In vitro diagnostic blood test Clinical and biological data were collected during a unique visit. Thus, participation ended after completing the interview and blood sample collection. Blood samples are collected in two PAXgene ® RNA collection tubes (2.5 mL each), shipped to the central laboratory and stored at −80 °C until analysis. Samples analyses were performed in batches. The extracted RNA from the blood and remaining blood samples were stored at −80 °C. The assay protocol was previously described in Salvetat et al 2024. To perform the test, blood samples are collected using standard procedures, and biological analyses were conducted by the Synlab laboratory accredited for RNA sequencing (in accordance with EN UN ISO 15189:2023). The panel of pre-mRNA sequences includes GAB2 (growth factor receptor bound protein 2-associated protein 2); IFNAR1 (interferon alpha/beta receptor 1); LYN (tyrosine-protein kinase Lyn); MDM2 (E3 ubiquitin-protein ligase Mdm2); PRKCB (protein kinase C beta type); IL17RA (interleukin 17 receptor A), PTPRC (protein tyrosine phosphatase receptor type C, also called CD45 antigen) and ZNF267 (zinc finger protein 267)( 25 ). The PAXgene® blood RNA tubes with collected blood samples were stored upright at room temperature (18−25°C) for 2 hours before being transferred to freezer (−20°C). The samples were stored locally at -20°C and sent on dry ice every two months to the Genetic Unit of SYNLAB Italia laboratory using time- and temperature-controlled transport. After reception, blood samples were kept at -80°C at SYNLAB laboratory upon subsequent analysis, which was performed in batch of randomly distributed patients. The biological protocol for Targeted Next Generation Sequencing was previously detailed( 25 ). Briefly, RNA was extracted and qualified before PCR, then the PCR product amplification was purified and quantified before indexing and pooling into a final library. The resulting library was denatured, loaded and sequenced at standard concentrations on an Illumina NextSeq 500/550 sequencer. 2.5 EDIT-B project web service platform and analysis EDIT-B project web platform Sequencing data were converted from BCL to FASTQ format. FASTQ files and corresponding metadata (age, sex, treatment, alcohol and tobacco use) were uploaded to the EDIT-B project web secure platform, where each patient was processed individually on a Health Data Hosting (HDS)–certified server. The HDS regulatory certification, based on the ISO/IEC 27001 standard, allows to demonstrate the commitment to protecting personal health data. For each patient and algorithm, the EDIT-B project web secure platform produced a standardized report including the prediction result, laboratory and sample identifiers, quality control metrics (sequencing and alignment), and algorithm specifications. Bioinformatics protocol Each patient data was analyzed by using ALCEDIAG proprietary bioinformatics( 25 ). A minimal sequencing depth of 10,000 reads for each sample was mandatory for further analysis. A pretreatment step and quality assessment (short reads removal, trimming, quality filter) was performed using fastp version 0.20.1( 40 ) and FastQC softwares (version 0.11.7, https://github.com/s-andrews/FastQC/ ). Alignment of the processed reads to the reference human genome sequence GRCh38 was performed using bowtie2 (version 2.2.9)( 41 ). The mapped reads were filtered considering quality criteria, using SAMtools software (version 1.11)( 42 ). Finally, SAMtools mpileup was used for SVN calling and finding edited positions in the alignment was done by using an in-house script to count the number of different nucleotides in each genomic location and calculate the editing biomarkers of interest. These biomarkers were combined with the metadata information for each one of the three algorithms, A154, A284 and A84 using ExtraTrees (ET) or extremely randomized trees algorithm( 43 ). For this, we used a grid learning approach stating a certain maximum hyper-parameter set (meanly, number of trees in the forest, maximum depth of the tree, minimum number of samples required to split an internal node). The subtrees were trained with a classical cross-validation and randomly repeated 500 times, using a training set to construct the model and a test set to validate it( 25 ). The implementation was done using the R ranger package (version 0.13.1)( 44 ) and R caret package (version 6.0–90)( 45 ). Candidate Algorithms A154, A284 and A84 In a recent publication( 25 ), we developed and validated a diagnostic algorithm based on the ExtraTree method, trained on a combination of RNA editing variants and clinical data. This model incorporated 22 RNA editing biomarkers, together with demographic (age, sex) and clinical variables, consisting in psychiatric treatments (antiepileptics, anxiolytics, hypnotics/sedatives, antidepressants—distinguishing SSRIs from other classes— antipsychotics, and lithium), as well as alcohol and tobacco use. This previously validated model is referred to as Algorithm A284 in the present study. This approach was extended by evaluating two additional models, designated A154 and A84, which were developed using the same cohorts and identical methodological framework (including preprocessing steps, internal parameter grid, and the ExtraTree algorithm). These additional models were designed to explore whether alternative hyperparameters and/or feature sets could yield improved classification performance, and to benchmark the robustness of A284 against comparable approaches derived from the same data. Algorithm A154: Built on the same 22 RNA editing biomarkers as A284, but with modified hyper-parameter values. In contrast to A284, treatment features in A154 were simplified, with SSRIs integrated into the broader antidepressant class and lithium grouped within the antipsychotic class. Algorithm A84: Constructed using an alternative panel of RNA editing biomarkers (n = 19), and with the same clinical features as A154. A84 also applied distinct hyper-parameter values compared to both A154 and A284. All three algorithms were developed prior to the initiation of EDIT-B project and were applied to the present analysis dataset (n = 393), following the same methodological procedures previously described( 25 ). 2.6 Data collection and management Eligible participants completed standardized assessments—MADRS, YMRS, MINI, and EQ-5D( 46 )—at pre- and post-diagnosis. Additional data included sociodemographics, clinical/psychiatric history (personal/family), illness characteristics (type, duration, onset age), comorbidities, substance use, and current medications. Data were recorded in an electronic case report form (e-CRF) by investigators and/or coordinators, with automated online checks for data integrity. A clinical research assistant (CRA) monitored and validated data throughout the study. After completion, final datasets were extracted, locked, and certified. A separate database for EDIT-B results (BD/MDD classification) was locked post-analysis. Participants providing informed consent and not withdrawing were included. Descriptive statistics were computed for the sample. A quality assurance system monitored study conduct and data integrity. Algorithmic results were blinded to investigators until study completion. All procedures complied with the European General Data Protection Regulation (GDPR). Participant data were pseudonymized, with encryption keys accessible only to the principal investigator and site coordinator. 2.7 Data analysis Outcomes The primary aim was to identify among three RNA biomarker algorithms the best in terms of external validation of the diagnostic performance for differentiating a current depressive episode as part of a BD from a current depressive episode as part of a MDD based on sensitivity, specificity, accuracy, and AUC-ROC. Diagnostic classification by clinicians (MINI-based) served as the gold standard. Statistical Analysis Performance of the tests was estimated by the sensitivity (i.e. the probability that a test will indicate ’bipolar disorder’ among those with this disease), specificity (i.e. the fraction of those without BD who will have a negative test result) and overall accuracy. Receiver Operating Characteristic (ROC) curves were generated to assess binary classification performance using pROC R package (v1.18.5). DeLong’s test( 47 ) compared AUC-ROC values between algorithms, while McNemar’s test( 48 ); R package DTComPair v1.2.6] evaluated pairwise differences in sensitivity and specificity. An optimized cut-off for positive predictive value (PPV) was determined using the cutpointr R package (v1.2.1)( 49 ) in a previously published cohort( 25 ). This threshold was selected to prioritize clinical utility by minimizing false positives, thereby reducing the risk of misdiagnosing BD. Descriptive statistics summarized demographic and clinical characteristics, including socio-demographics and clinical features, assessment scores (MADRS, YMRS, EQ-5D-5L, MINI), medication use (coded via WHO ATC classes) and substance use. 3. RESULTS 3.1 Patients demographic and clinical characteristics Patients were matched for sex, age, and severity of depression based on MADRS score, with no significant between-group differences observed. The proportion of female patients was similar between the major depressive disorder (MDD) and bipolar disorder (BD) groups (61.8% vs. 61.3%, respectively; p-value = 1). Mean age did not differ significantly between groups (47.0 years for MDD vs. 45.0 years for BD; p = 0.2). Patients were largely homogeneous in terms of race/ethnicity, with 90.8% of participants identifying as White/Caucasian. No significant differences in racial/ethnic distribution were observed between patients with MDD and those with BD. Body mass index (BMI) was comparable across groups. Depression severity was assessed by the MADRS score, which was not statistically significantly different (p-value = 0.9) between MDD patients (mean score of 32.1) and BD patients (mean score of 32.2). In contrast, BD patients had statistically significantly higher YMRS scores than major depressive disorder patients (p-value < 0.001), but its YMRS score stays very low in the BD group (YMRS score = 1.1 for the MDD vs 1.9 for the BD). These results confirm that the BD patients were not in a manic phase ( Table 1 ). Regarding psychotropic treatments, approximately 79 % of the patients were treated with antidepressants, 63 % with antipsychotics, 35 % of patients received anxiolytics, while 21 % of patients received hypnotics and sedatives and 35 % received antiepileptics (see Table 2 ). Smoking was reported by 37% of patients, and alcohol abuse by 8%, with no significant between-group differences (p-value = 1) ( Table 1 ). View this table: View inline View popup Download powerpoint Table 2. Pairwise McNemar’s tests of sensitivity and specificity across the three algorithms (N=393). View this table: View inline View popup Download powerpoint Table 3. Diagnostic performances for the Algorithm A284, Analytical Set (N = 393). Abbreviations: AUC = Area Under Curve; CI95% = Confidence Interval 95%; PPV: positive predictive value; NPV: negative predictive value. The mean duration of illness was comparable between the two groups (approximately 10–11 years, see Supplementary Table 2). Although the onset of disease and the first depressive episode tended to occur earlier in patients with BD, the difference was not statistically significant. Both groups exhibited considerable variability (SD) in illness duration and age at onset. Among the BD patients, the first depressive episode typically preceded the first manic or hypomanic episode. Hypomanic episodes were more frequently reported than manic episodes, and BD patients experienced a higher number of depressive episodes compared to MDD patients. Depressive polarity at onset was predominant in BD patients (88.4%). Psychotic symptoms were present in approximately one-third of BD patients, compared with fewer than 10% of MDD patients. A lifetime history among first or second generations of affective disorders was reported by 75% of BD patients and 50% of MDD patients, while a family history of psychiatric illness was more common in BD patients (81.3%) than in MDD patients (60.5%). A family history among first or second generations of completed suicide was reported by approximately 20% of patients in both groups (Supplementary Table 2: Disease Characteristics). According to the Mini International Neuropsychiatric Interview (MINI), all participants fulfilled criteria for an ongoing MDE of at least two weeks’ duration. The majority also reported a history of past and/or recurrent MDEs. None of the MDD patients endorsed a lifetime manic or hypomanic episode. Among BD patients, 40.0% reported a lifetime manic episode and 63.9% reported a lifetime hypomanic episode. Suicidal ideation or behavior was common in both groups, with 20–30% reporting a recent suicide attempt. The structured interview also revealed the following psychiatric comorbidities: panic disorder (19.1% overall; 26.5% in BD vs. 14.3% in MDD), generalized anxiety disorder (15.0%, similar across groups), psychotic disorders (3.8%, predominantly in BD I), alcohol use disorder (7.6%), substance use disorders (4.3%), obsessive-compulsive disorder (8.4%), and eating disorders (2.0%) (Supplementary Table 3: MINI). 3.2 Comparison of the AUC: A284 vs A154 vs A84 The three algorithms were evaluated and compared using area under the receiver operating characteristic curve (AUC-ROC) analysis ( Figure 3 ). Download figure Open in new tab Figure 3. Comparison of the ROC curves obtained with the 3 algorithms: A) Algorithm A284; B) Algorithm A154; C) Algorithm A84. Legend. Dark blue line: Empirical Area Under ROC Curve (AUC); dotted light blue line: Generalized Area Under ROC Curve. The empirical ROC curves gave an AUC-ROC of 0.873 (CI 95%: [0.837 - 0.909]) for Algorithm A284, 0.834 (CI 95%: [0.793 - 0.875]) for Algorithm A154 and 0.741 (CI 95%: [0.690 - 0.792]) for Algorithm A84. For the algorithms A284 and A154, the generalized ROC curves closely align with the empirical ROC curves, indicating that their estimated performances are consistent with observed data. In contrast, the generalized ROC curve for Algorithm A84 only moderately aligned with its empirical counterpart, suggesting that its estimated performance partially reflects the actual observed data. Pairwise comparisons of the three algorithms were conducted using the McNemar test to assess sensitivity and specificity ( Table 2 ). Sensitivity did not differ significantly among the three signatures. However, specificity comparisons revealed statistically significant differences: A284 vs. A154 (p < 0.05), A154 vs. A84 (p < 0.001), and A284 vs. A84 (p < 0.001), with the most pronounced differences observed between Algorithm A84 and the other two algorithms. 3.3 Comparison of A284 AUC ROC with previous results The empirical ROC curve of Algorithm A284, which is the same as that previously published( 25 ), was compared to prior results (see Figure 4 ). The 95% confidence intervals (CIs) for the AUC-ROC of A284 in this study overlapped with those previously reported, confirming statistical consistency. No significant difference in performance was observed between the two ROC curves (DeLong’s test, p = 0.33). Overall, these findings reaffirm that Algorithm A284 consistently performs well in diagnosis, outperforming all other candidate algorithms evaluated in this study. Download figure Open in new tab Figure 4. Comparison of the ROC curves and confidence intervals of the algorithms A284 vs previous study 25 . Light blue line : previous study ROC curve (Salvetat et al, 2024); Dark blue line: Algorithm A284 ROC Curve; P-value = p-value of the DeLong’s Test. 3.4 Algorithm A284 diagnostic performance The diagnostic performances for the algorithm A284 are shown in Table 4. In this study (N=393), Algorithm A284 achieved a sensitivity of 82.6% (CI 95%: 75.7% - 88.2%), a specificity of 80.3% (CI 95%: 74.6% - 85.1%), a negative predictive value (NPV) of 87.6% (CI 95%: 65.9% - 79.6%), a positive predictive value of 73.1% (CI 95%: 82.5% - 91.7%) and an overall accuracy of 81.2% (CI 95%: 77.0% - 84.9%). These results demonstrate that Algorithm A284 provides a robust separation between patients with MDD and those with BD. 3.5 BD reference centers analysis Among the 4 recruiting centers involved in this study, 3 were BD reference centers (Center 001, Spain; Center 002, Denmark; and Center 003, France). An exploratory analysis was conducted to assess the AUC-ROC curves across the three BD reference centers, comparing the performance of the three algorithms (Supplementary Figure 1). Overall, Algorithm A284 demonstrated the highest performance across all three centers (AUCs of 0.898 (Center 001), 0.887 (Center 002), and 0.893 (Center 003) respectively). Notably, AUC ROC curves at the BD reference centers were superior to the non-BD reference center (Center 004, Spain; AUCs for A284: 0.844). When combining data from the three BD reference centers (N=295), Algorithm A284 achieved an empirical AUC of 0.891 and a generalized AUC of 0.894 (Supplementary Figure 2A). This performance did not differ significantly from that of the previously described( 25 ) ( p = 0.7, Supplementary Figure 2B). The diagnostic performances of Algorithm A284 were further compared across two populations: (i) BD reference centers only (Centers 001, 002, and 003; Supplementary Table 3) and (ii) all centers combined (Centers 001–004). The BD reference center population demonstrated superior AUC (0.891, 95% CI: 0.854–0.929) as compared to the full population (0.873, 95% CI: 0.837–0.909), as well as higher overall accuracy (84.1% vs. 81.2%). Interestingly, when considering all centers, sensitivity and negative predictive value (NPV) were slightly higher than for BD reference centers only (sensitivity: 82.6% vs. 81.5%; NPV: 87.6% vs. 84.7%). By contrast, the BD reference center population yielded superior specificity (86.3% vs. 80.3%) and positive predictive value (PPV) (83.3% vs. 73.1%), consistent with its higher overall accuracy. 3.6 Algorithm A284 across sex, age and BD subtypes The diagnostic performance of the A284 algorithm was evaluated across key clinical parameters including age, sex and BD subtypes. Sex-based subgroup analyses revealed no significant differences between MDD and BD distribution (Chi² test, p = 1.0) or comparable diagnostic performance for males (AUC = 0.888) and females (AUC = 0.863; DeLong’s test, p = 0.49; Supplementary Figure 3A). Three age intervals have been defined, following their distribution, to compare MDD and BD performances: 18 to 35 years, 36 to 55 years, and finally, more than 55 years. Across all age strata (18–35, 36–55, and >55 years), Algorithm A284 demonstrated robust diagnostic performance with AUC values ranging from 0.865 to 0.894 and no significant differences observed between groups (DeLong’s test p = 0.69, 0.80, and 0.53 for pairwise comparisons; two-way Anova p = 0.92; Supplementary Figure 3B). Regarding BD subtypes (BD type I and BD type II), a significant difference in AUC-ROC was observed between subtypes (DeLong’s test, p = 0.04), with BD type I exhibiting higher discriminatory power (AUC = 0.913) as compared to BD type II (AUC = 0.849). Despite this difference, both subtypes demonstrated high diagnostic accuracy (>80%): 81.1% for BD Type I and 80.5% for BD type II (data not shown) indicating robust overall performance of the diagnostic tool across BD subtypes (Supplementary Figure 3C). 4. DISCUSSION This study demonstrates that the A-to-I RNA-editing based test, as an innovative and robust complement to clinical evaluation, represents a major advance in the diagnostic workup of mood disorders. By significantly improving the accuracy of the differential diagnosis between MDD and BD, this cutting-edge approach could contribute to reduce the diagnostic delay for BD, enabling an earlier management of patients with mood disorders. This earlier diagnostic in clinical management could enable more effective treatment decisions, reduce the risk of antidepressant-induced mania( 50 ) and ultimately lower hospitalization rates and associated healthcare costs. Ultimately and beyond individual benefits, the implementation of the test in clinical routine could reduce the economic burden and optimize healthcare resource allocation. Taken together, these results position the test as a clinically actionable innovation with direct translational value for personalized and precision psychiatry. Our previous study demonstrated that a panel of blood A-to-I RNA editing biomarkers can differentiate MDD from BD and show a validation of this approach in an independent multicenter cohort in Switzerland( 25 ). In the present study, we extend this work by evaluating three diagnostic algorithms in a third, large and independent European multicenter cohort spanning three countries France, Spain and Denmark, with the aim of benchmarking performance and assessing generalizability. Importantly, Algorithm A284—previously validated( 25 ) emerged as the most robust model, combining high discriminative performance (accuracy, sensitivity and specificity >80%) with strong reproducibility across diverse clinical settings. When applied to 393 patients from four centers in three countries, A284 closely replicated the results of the original validation study( 25 ), underscoring its stability across independent cohorts. Moreover, its performance was even greater when the set of analysis was restricted to patients clinically diagnosed at BD reference centers (N = 295), supporting both its clinical reliability and its potential for broad generalizability in real-world practice. There is a slight difference in concordance between the specialized reference centers for BD and the general psychiatry center. Indeed, a lower Positive Predictive Value (PPV) indicates a higher proportion of false positives in this center, which slightly reduces the overall accuracy of the model. It is crucial that the diagnosis established by clinicians be as reliable as possible, since this test compares the clinical diagnosis with the RNA editing-based IVD test combined with the three algorithms evaluated. This finding highlights the usefulness of such a tool as a diagnostic aid, particularly in general psychiatry centers. Regarding the study population, the results exhibited largely comparable baseline clinical characteristics between participants with MDD and BD, including age, sex, race, BMI, psychiatric treatment, suicidality, suicide behavior, and most comorbidities, with the exception of a higher prevalence of panic disorders in BD. This consistency confirms that the recruitment process rigorously adhered to predefined inclusion criteria, ensuring balanced group comparisons. Moreover, while the mean duration of illness was comparable across groups, BD tended to manifest earlier, with the first depressive episode typically preceding the onset of mania or hypomania. In line with this, BD patients experienced a higher lifetime number of depressive episodes compared with MDD patients, reinforcing the recurrent and disabling nature of bipolar depression( 51 ). Family history patterns also revealed notable differences. A psychiatric family history was more prevalent in BD than in MDD, confirming the strong heritable component of bipolar disorder, reinforcing its genetic underpinnings. Taken together, these findings underscore the complexity of mood disorders and the challenges of differentiating BD from MDD based solely on clinical features during depressive episodes. Expertise of psychiatrists is essential for the accurate evaluation of bipolar disorder, but based on clinical features, the diagnostic distinction of a major depressive episode as part of BD from a major depressive episode as part of MDD remains relatively subjective, increasing the time to a correct diagnosis( 3 ). Current standardized screening tools can show an insufficient discriminatory power: highly trained psychiatrists agreed on the diagnosis between 4 and 15% of the time for trials for the Diagnostic and Statistical Manual of Mental Disorders (DSM-5), which demonstrated an intra class kappa value of 0.28( 52 ). Although, the Mood Disorder Questionnaire (MDQ) has demonstrated a sensitivity slightly above 70% in outpatient psychiatric clinics( 53 ), as well as good accuracy for screening BD in primary care( 54 ), some studies indicate that in patients with bipolar spectrum illness, the sensitivity reaches 70% only for BD type I( 55 ). Meanwhile, scales such as the 32-item Hypomania symptom check-list, first revision (HCI-32-R1), have shown utility in detecting hypomanic symptoms( 56 ), which can further aid in identifying at risk patients. The EDIT-B panel (strongly associated to central nervous system function, mental disorders, and immune-related diseases( 25 )) highlights the potential of molecular signatures to elucidate the biological underpinnings of depression. Meanwhile, the integration of RNA sequencing and machine learning is emerging as a powerful approach to enhance diagnostic precision in medicine, enabling the identification of robust disease-specific molecular signatures with high precision. This paradigm is already well established in oncology, where FDA- and CE-approved assays such as Guardant360 (oncology)( 57 ), MammaPrint (breast cancer)( 58 ), and Prolaris (prostate cancer)( 59 ) have demonstrated the clinical and economic value of such approaches. These tools not only refine diagnosis and stratify patients for personalized treatment but also minimize unnecessary interventions and optimize healthcare resources. This approach highlights the transformative potential of AI-driven genomics and epigenomics in improving clinical outcomes, including for complex disorders in psychiatry, where epitranscriptomic signatures such as the EDIT-B gene panel show great promise. Additionally, the identification of genes associated with the diagnosis an allow advances in the understanding of the pathophysiology of depression, especially in understanding the participation of the monoamine hypothesis, hypothalamic-pituitary-adrenal axis changes, neuroplasticity and neurogenesis, epigenetics and inflammation 61 . Limitations The present study has several limitations that must be considered. First, the majority of participants were of Caucasian origin. While this reflects the demographic profile of the recruitment centers, it may limit the generalizability of our findings to more diverse populations. Second, we did not account for the low severity depression states and the different bipolar sub-classifications (i.e. BD1 vs. BD2) in the algorithm A284. Indeed, diagnostic performance differed between BD1 and BD2 disorders in distinguishing MDD from BD. Nonetheless, both subtypes achieved high diagnostic accuracy (>80%), underscoring the robustness of the assay across BD classifications. Moreover, the higher prevalence of psychotic features in BD I compared with BD II and MDD may further limit the generalizability of our findings. Third, the mean age of the study population was relatively high (45-47 years old) compared to the typical age at which the assay would be implemented in clinical practice, which may potentially limit the extrapolation of these findings to younger patient populations. However, a subgroup analyses stratified by age demonstrated consistent diagnostic performance of the assay, with no significant differences in accuracy, sensitivity, or specificity observed across study sites. Finally, the study was prospective but cross-sectional, although based on patients with a proven BD or MDD trajectory; a longitudinal study with first episode major depression would provide the greatest evidence in favor of the predictive value of the test( 60 ). In sum, this study provides further evidence and replication of then validity and effectiveness of an AI-driven RNA-editing biological test for the objective diagnosis of bipolar disorder in depressed patients. This implies a great advance in the context of precision psychiatry( 61 ) and may help to reduce misdiagnosis and time spent without proper treatment. Abbreviations BD Bipolar disorder MDD Major depressive disorder DNA Deoxyribonucleic acid RNA Ribonucleic acid AUC-ROC Area under the receiver operating characteristic curve MDE Major depressive episode MINI Mini-International Neuropsychiatric Interview DSM-5 Diagnostic and Statistical Manual of Mental Disorders 5 MADRS Montgomery–Åsberg Depression Rating Scale YMRS Young Mania Rating Scale EQ-5D European Quality of Life 5 Dimensions Questionnaire MDQ Mood Disorder Questionnaire HCI-32-R1 32-Item Hypomania symptom check-list, first revision CRA Clinical research assistant ADHD Attention-Deficit/Hyperactivity Disorder Author contributions E.V, DW: lead the study, secured funding. E.V, DW, MF: conceived, designed the study LVK, CH, JMH, E.V: diagnosed patient, supervised and collected data. DW, NS, FCR: draft the initial manuscript. NS, FCR: prepared figures and tables and conducted data analyses and interpretation. LVK, CH, JMH, E.V: reviewed and edited the manuscript. CC: conceived and designed biological protocols, provided support in conducting the study. DV: clinical monitoring and managed the study. LWB, GR: performed analyses. AMM, JZ, AGP, MGC, MV: diagnosed patient and collected data. Data Availability All data produced in the present study are available upon reasonable request to the authors Funding This work was supported by EIT Health [220628/230125]. Ethics approval and consent to participate Ethical approval has been obtained in France by the research ethics committee Ile-de-France X (ID 2021-A-03194-37), in Spain by the research ethics committees of the Hospital Clinic of Barcelona (ID HCB/2022/0042) and the Sant Joan de Déu Foundation (PS-05-22) and in Denmark by the De Videnskabsetiske Komiteer for Region Hovedstaden (the Scientific Ethics Committees for the Capital Region of Denmark) (ID H-22010899). Disclosures DW is the CSO of Alcediag. NS, CC, FCR, DV are employed by Alcediag. LWB, GR, MF are employed by Synlab. EV has received grants and served as consultant, advisor or CME speaker for the following entities: AB-Biotics, Abbott, AbbVie, Adamed, Adium, Alcediag, Angelini, Biogen, Beckley- Psytech, Biohaven, Boehringer-Ingelheim, Casen-Recordati, Celon Pharma, Compass, Dainippon Sumitomo Pharma, Esteve, Ethypharm, Ferrer, Gedeon Richter, GH Research, Glaxo-Smith Kline, HMNC, Intra-Cellular therapies, Idorsia, Johnson & Johnson, Lundbeck, Luye Pharma, Medincell, Merck, Mitsubishi Tanabe Pharma, Newron, Novartis, Organon, Orion Corporation, Otsuka, Roche, Rovi, Sage, Sanofi-Aventis, Sunovion, Takeda, Teva, and Viatris, outside the submitted work. LVK has within recent three years been a consultant for Teva and Lundbeck. CH, JMH has no conflict of interest. Acknowledgements The authors would like to acknowledge the contributions of Aixial for their support in conducting the study. References 1. ↵ Dagani J , Signorini G , Nielssen O , Bani M , Pastore A , Girolamo G de , Large M ( 2017 ): Meta-analysis of the Interval between the Onset and Management of Bipolar Disorder . Can J Psychiatry 62 : 247 – 258 . OpenUrl CrossRef PubMed 2. Fritz K , Russell AMT , Allwang C , Kuiper S , Lampe L , Malhi GS ( 2017 ): Is a delay in the diagnosis of bipolar disorder inevitable? Bipolar Disord 19 : 396 – 400 . OpenUrl CrossRef 3. ↵ Hirschfeld RM ( 2014 ) : Differential diagnosis of bipolar disorder and major depressive disorder . Journal of Affective Disorders 169 : S12 – S16 . OpenUrl CrossRef PubMed 4. Nierenberg AA , Agustini B , Köhler-Forsberg O , Cusin C , Katz D , Sylvia LG , et al. ( 2023 ) : Diagnosis and Treatment of Bipolar Disorder: A Review . JAMA 330 : 1370 . OpenUrl CrossRef PubMed 5. ↵ Scott J , Graham A , Yung A , Morgan C , Bellivier F , Etain B ( 2022 ): A systematic review and meta-analysis of delayed help-seeking, delayed diagnosis and duration of untreated illness in bipolar disorders . Acta Psychiatr Scand 146 : 389 – 405 . OpenUrl CrossRef PubMed 6. ↵ Altamura AC , Buoli M , Caldiroli A , Caron L , Cumerlato Melter C , Dobrea C , et al. ( 2015 ) : Misdiagnosis, duration of untreated illness (DUI) and outcome in bipolar patients with psychotic symptoms: A naturalistic study . Journal of Affective Disorders 182 : 70 – 75 . OpenUrl CrossRef PubMed 7. Altamura AC , Dell’Osso B , Berlin HA , Buoli M , Bassetti R , Mundo E ( 2010 ): Duration of untreated illness and suicide in bipolar disorder: a naturalistic study . Eur Arch Psychiatry Clin Neurosci 260 : 385 – 391 . OpenUrl CrossRef PubMed 8. Bauer M , Andreassen OA , Geddes JR , Vedel Kessing L , Lewitzka U , Schulze TG , Vieta E ( 2018 ): Areas of uncertainties and unmet needs in bipolar disorders: clinical and research perspectives . The Lancet Psychiatry 5 : 930 – 939 . OpenUrl PubMed 9. Macaluso M , Preskorn SH Beyer JL ( 2019 ): The Use of Antidepressants in Bipolar Depression . In: Macaluso M , Preskorn SH , editors. Antidepressants: From Biogenic Amines to New Mechanisms of Action . Cham : Springer International Publishing , pp 415 – 442 . 10. Kessing LV , Andersen PK ( 2017 ): Evidence for clinical progression of unipolar and bipolar disorders . Acta Psychiatr Scand 135 : 51 – 64 . OpenUrl CrossRef PubMed 11. ↵ Stang PE , Frank C , Kalsekar A , Yood MU , Wells K , Burch S ( 2006 ): The clinical history and costs associated with delayed diagnosis of bipolar disorder . MedGenMed 8 : 18 . OpenUrl PubMed 12. ↵ Carvalho AF , Firth J , Vieta E ( 2020 ): Bipolar Disorder ((A. H. Ropper, editor)) . N Engl J Med 383 : 58 – 66 . OpenUrl CrossRef PubMed 13. ↵ Rowland T , Perry BI , Upthegrove R , Barnes N , Chatterjee J , Gallacher D , Marwaha S ( 2018 ): Neurotrophins, cytokines, oxidative stress mediators and mood state in bipolar disorder: systematic review and meta-analyses . Br J Psychiatry 213 : 514 – 525 . OpenUrl PubMed 14. Sowa-Kućma M , Styczeń K , Siwek M , Misztak P , Nowak RJ , Dudek D , et al. ( 2018 ): Lipid Peroxidation and Immune Biomarkers Are Associated with Major Depression and Its Phenotypes, Including Treatment-Resistant Depression and Melancholia . Neurotox Res 33 : 448 – 460 . OpenUrl PubMed 15. ↵ Wollenhaupt-Aguiar B , Librenza-Garcia D , Bristot G , Przybylski L , Stertz L , Kubiachi Burque R , et al. ( 2020 ) : Differential biomarker signatures in unipolar and bipolar depression: A machine learning approach . Aust N Z J Psychiatry 54 : 393 – 401 . OpenUrl PubMed 16. ↵ Cui P , Han D-Y , Yang C-H ( 2025 ): Early diagnosis of bipolar disorder . World J Psychiatry 15 . doi: 10.5498/wjp.v15.i8.106125 OpenUrl CrossRef 17. ↵ He D , Yang J , Wang Z , Dong X ( 2025 ): Biomarkers of bipolar disorder in omics and neuroimaging . Journal of Pharmaceutical Analysis 101264 . 18. ↵ Ansell BRE , Thomas SN , Bonelli R , Munro JE , Freytag S , Bahlo M ( 2021 ): A survey of RNA editing at single-cell resolution links interneurons to schizophrenia and autism . RNA 27 : 1482 – 1496 . OpenUrl Abstract / FREE Full Text 19. Barbon A , Magri C ( 2020 ): RNA Editing and Modifications in Mood Disorders . Genes 11 : 872 . OpenUrl 20. Chimienti F , Cavarec L , Vincent L , Salvetat N , Arango V , Underwood MD , et al. ( 2019 ): Brain region-specific alterations of RNA editing in PDE8A mRNA in suicide decedents . Transl Psychiatry 9 : 91 . OpenUrl PubMed 21. Di Narzo AF , Kozlenkov A , Roussos P , Hao K , Hurd Y , Lewis DA , et al. ( 2014 ): A unique gene expression signature associated with serotonin 2C receptor RNA editing in the prefrontal cortex and altered in suicide . Human Molecular Genetics 23 : 4801 – 4813 . OpenUrl CrossRef PubMed 22. Gurevich I , Englander MT , Adlersberg M , Siegal NB , Schmauss C ( 2002 ): Modulation of Serotonin 2C Receptor Editing by Sustained Changes in Serotonergic Neurotransmission . J Neurosci 22 : 10529 – 10532 . OpenUrl Abstract / FREE Full Text 23. Hayashi MAF , Salvetat N , Cayzac C , Checa-Robles FJ , Dubuc B , Mereuze S , et al. ( 2023 ): Euthymic and depressed bipolar patients are characterized by different RNA editing patterns in blood . Psychiatry Research 328 : 115422 . OpenUrl PubMed 24. Maas S , Kawahara Y , Tamburro KM , Nishikura K ( 2006 ): A-to-I RNA Editing and Human Disease . RNA Biology 3 : 1 – 9 . OpenUrl PubMed 25. ↵ Salvetat N , Checa-Robles FJ , Delacrétaz A , Cayzac C , Dubuc B , Vetter D , et al. ( 2024 ): AI algorithm combined with RNA editing-based blood biomarkers to discriminate bipolar from major depressive disorders in an external validation multicentric cohort . Journal of Affective Disorders 356 : 385 – 393 . OpenUrl PubMed 26. Salvetat N , Chimienti F , Cayzac C , Dubuc B , Checa-Robles F , Dupre P , et al. ( 2021 ): Phosphodiesterase 8A to discriminate in blood samples depressed patients and suicide attempters from healthy controls based on A-to-I RNA editing modifications . Translational Psychiatry 11 . doi: 10.1038/s41398-021-01377-9 OpenUrl CrossRef 27. Salvetat N , Van der Laan S , Vire B , Chimienti F , Cleophax S , Bronowicki JP , et al. ( 2019 ): RNA editing blood biomarkers for predicting mood alterations in HCV patients . J Neurovirol 25 : 825 – 836 . OpenUrl CrossRef PubMed 28. Silberberg G , Lundin D , Navon R , Öhman M ( 2012 ): Deregulation of the A-to-I RNA editing mechanism in psychiatric disorders . Human Molecular Genetics 21 : 311 – 321 . OpenUrl CrossRef PubMed Web of Science 29. Slotkin W , Nishikura K ( 2013 ): Adenosine-to-inosine RNA editing and human disease . Genome Med 5 : 105 . OpenUrl CrossRef PubMed 30. Weissmann D , van der Laan S , Underwood MD , Salvetat N , Cavarec L , Vincent L , et al. ( 2016 ): Region-specific alterations of A-to-I RNA editing of serotonin 2c receptor in the cortex of suicides with major depression . Transl Psychiatry 6 : e878 – e878 . OpenUrl PubMed 31. ↵ Yang Y , Okada S , Sakurai M ( 2021 ): Adenosine-to-inosine RNA editing in neurological development and disease . RNA Biology 18 : 999 – 1013 . OpenUrl PubMed 32. ↵ Miranda-Mendizabal A , Vetter D , Zambrano J , Zarp J , Chavarría V , Giménez-Palomo A , et al. ( 2025 ): RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: protocol of the EDIT-B study . Ann Gen Psychiatry 24 : 7 . OpenUrl PubMed 33. ↵ Chan A-W , Tetzlaff JM , Gotzsche PC , Altman DG , Mann H , Berlin JA , et al. ( 2013 ): SPIRIT 2013 explanation and elaboration: guidance for protocols of clinical trials . BMJ 346 : e7586 – e7586 . OpenUrl Abstract / FREE Full Text 34. ↵ Fusar-Poli P , Manchia M , Koutsouleris N , Leslie D , Woopen C , Calkins ME , et al. ( 2022 ) : Ethical considerations for precision psychiatry: A roadmap for research and clinical practice . European Neuropsychopharmacology 63 : 17 – 34 . OpenUrl CrossRef PubMed 35. ↵ Buderer NMF ( 1996 ) : Statistical Methodology: I. Incorporating the Prevalence of Disease into the Sample Size Calculation for Sensitivity and Specificity . Academic Emergency Medicine 3 : 895 – 900 . OpenUrl CrossRef PubMed Web of Science 36. ↵ Jones SR , Carley S , Harrison M ( 2003 ): An introduction to power and sample size estimation . Emerg Med J 20 : 453 – 458 . OpenUrl Abstract / FREE Full Text 37. ↵ Sheehan DV , Lecrubier Y , Sheehan KH , Amorim P , Janavs J , Weiller E , et al. ( 1998 ): The Mini-International Neuropsychiatric Interview (M.I.N.I.): the development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10 . J Clin Psychiatry 59 . 38. ↵ Montgomery SA , Åsberg M ( 1979 ): A New Depression Scale Designed to be Sensitive to Change . The British Journal of Psychiatry 134 : 382 – 389 . OpenUrl Abstract / FREE Full Text 39. ↵ Young RC , Biggs JT , Ziegler VE , Meyer DA ( 1978 ) : A Rating Scale for Mania: Reliability, Validity and Sensitivity . The British Journal of Psychiatry 133 : 429 – 435 . OpenUrl Abstract / FREE Full Text 40. ↵ Chen S , Zhou Y , Chen Y , Gu J ( 2018 ): fastp: an ultra-fast all-in-one FASTQ preprocessor . Bioinformatics 34 : i884 – i890 . OpenUrl CrossRef PubMed 41. ↵ Langmead B , Salzberg SL ( 2012 ): Fast gapped-read alignment with Bowtie 2 . Nat Methods 9 : 357 – 359 . OpenUrl CrossRef PubMed Web of Science 42. ↵ Li H , Handsaker B , Wysoker A , Fennell T , Ruan J , Homer N , et al. ( 2009 ): The Sequence Alignment/Map format and SAMtools . Bioinformatics 25 : 2078 – 2079 . OpenUrl CrossRef PubMed Web of Science 43. ↵ Geurts P , Ernst D , Wehenkel L ( 2006 ): Extremely randomized trees . Mach Learn 63 : 3 – 42 . OpenUrl CrossRef 44. ↵ Wright MN , Ziegler A ( 2017 ): ranger: A Fast Implementation of Random Forests for High Dimensional Data in C++ and R . J Stat Soft 77 . doi: 10.18637/jss.v077.i01 OpenUrl CrossRef 45. ↵ Kuhn M ( 2008 ): Building Predictive Models in R Using the caret Package . J Stat Soft 28 . doi: 10.18637/jss.v028.i05 OpenUrl CrossRef PubMed 46. ↵ Janssen MF , Pickard AS , Golicki D , Gudex C , Niewada M , Scalone L , et al. ( 2013 ): Measurement properties of the EQ-5D-5L compared to the EQ-5D-3L across eight patient groups: a multi-country study . Qual Life Res 22 : 1717 – 1727 . OpenUrl CrossRef PubMed Web of Science 47. ↵ DeLong ER , DeLong DM , Clarke-Pearson DL ( 1988 ): Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach . Biometrics 44 : 837 – 845 . OpenUrl CrossRef PubMed Web of Science 48. ↵ Hawass NE ( 1997 ): Comparing the sensitivities and specificities of two diagnostic procedures performed on the same group of patients . The British Journal of Radiology 70 : 360 – 366 . OpenUrl Abstract 49. ↵ Thiele C , Hirschfeld G ( 2021 ): cutpointr1: Improved Estimation and Validation of Optimal Cutpoints in R . J Stat Soft 98 . doi: 10.18637/jss.v098.i11 OpenUrl CrossRef 50. ↵ Fico G , Vieta E ( 2024 ) : Antidepressant use in bipolar disorder: Shifting focus from ‘Whether’ to ‘Whom.’ European Neuropsychopharmacology 84 : 1 – 2 . OpenUrl PubMed 51. ↵ Oliva V , Fico G , De Prisco M , Gonda X , Rosa AR , Vieta E ( 2025 ): Bipolar disorders: an update on critical aspects . The Lancet Regional Health - Europe 48 : 101135 . OpenUrl PubMed 52. ↵ Freedman R , Lewis DA , Michels R , Pine DS , Schultz SK , Tamminga CA , et al. ( 2013 ) : The Initial Field Trials of DSM-5: New Blooms and Old Thorns . AJP 170 : 1 – 5 . OpenUrl 53. ↵ Hirschfeld RM , Williams JB , Spitzer RL , Calabrese JR , Flynn L , Keck PE , et al. ( 2000 ): Development and validation of a screening instrument for bipolar spectrum disorder: the Mood Disorder Questionnaire . Am J Psychiatry 157 : 1873 – 1875 . OpenUrl CrossRef PubMed Web of Science 54. ↵ Sasdelli A , Lia L , Luciano CC , Nespeca C , Berardi D , Menchetti M ( 2013 ) : Screening for Bipolar Disorder Symptoms in Depressed Primary Care Attenders: Comparison between Mood Disorder Questionnaire and Hypomania Checklist (HCL-32) . Psychiatry Journal 2013 : 1 – 9 . OpenUrl 55. ↵ Miller C ( 2004 ): Sensitivity and specificity of the Mood Disorder Questionnaire for detecting bipolar disorder . Journal of Affective Disorders 81 : 167 – 171 . OpenUrl CrossRef PubMed 56. ↵ Angst J , Meyer TD , Adolfsson R , Skeppar P , Carta M , Benazzi F , et al. ( 2010 ): Hypomania: a transcultural perspective . World Psychiatry 9 : 41 – 49 . OpenUrl PubMed Web of Science 57. ↵ Chung DC , Gray DM , Singh H , Issaka RB , Raymond VM , Eagle C , et al. ( 2024 ): A Cell-free DNA Blood-Based Test for Colorectal Cancer Screening . N Engl J Med 390 : 973 – 983 . OpenUrl CrossRef PubMed 58. ↵ Brandão M , Pondé N , Piccart-Gebhart M ( 2019 ): Mammaprint TM : a comprehensive review . Future Oncol 15 : 207 – 224 . OpenUrl PubMed 59. ↵ Hu JC , Tosoian JJ , Qi J , Kaye D , Johnson A , Linsell S , et al. ( 2018 ): Clinical Utility of Gene Expression Classifiers in Men With Newly Diagnosed Prostate Cancer . JCO Precision Oncology 1 – 15 . 60. ↵ Vieta E , De Prisco M ( 2024 ) : Cross-sectional studies: Is pressing the pause button worth it in research? European Neuropsychopharmacology 85 : 32 – 33 . OpenUrl CrossRef PubMed 61. ↵ Vieta E ( 2015 ) : Personalized medicine applied to mental health: Precision psychiatry . Revista de Psiquiatría y Salud Mental (English Edition) 8 : 117 – 118 . OpenUrl View the discussion thread. Back to top Previous Next Posted November 07, 2025. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: results of the EDIT-B consortium study Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: results of the EDIT-B consortium study D. Weissmann , N. Salvetat , C. Cayzac , F. Checa Robles , D. Vetter , L. Walczer-Baldinazzo , G. Ruggeri , M. Ferrari , A. Miranda-Mendizabal , J. Zambrano , J. Zarp , A. Giménez-Palomo , M. Valenti , JM. Haro , LV. Kessing , C. Henry , E. Vieta medRxiv 2025.11.05.25339583; doi: https://doi.org/10.1101/2025.11.05.25339583 Share This Article: Copy Citation Tools RNA editing-based biomarker blood test for the diagnosis of bipolar disorder: results of the EDIT-B consortium study D. Weissmann , N. Salvetat , C. Cayzac , F. Checa Robles , D. Vetter , L. Walczer-Baldinazzo , G. Ruggeri , M. Ferrari , A. Miranda-Mendizabal , J. Zambrano , J. Zarp , A. Giménez-Palomo , M. Valenti , JM. Haro , LV. Kessing , C. Henry , E. Vieta medRxiv 2025.11.05.25339583; doi: https://doi.org/10.1101/2025.11.05.25339583 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Psychiatry and Clinical Psychology Subject Areas All Articles Addiction Medicine (568) Allergy and Immunology (863) Anesthesia (300) Cardiovascular Medicine (4435) Dentistry and Oral Medicine (444) Dermatology (382) Emergency Medicine (608) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1509) Epidemiology (15229) Forensic Medicine (30) Gastroenterology (1124) Genetic and Genomic Medicine (6600) Geriatric Medicine (668) Health Economics (997) Health Informatics (4536) Health Policy (1368) Health Systems and Quality Improvement (1613) Hematology (541) HIV/AIDS (1264) Infectious Diseases (except HIV/AIDS) (15916) Intensive Care and Critical Care Medicine (1103) Medical Education (623) Medical Ethics (146) Nephrology (667) Neurology (6599) Nursing (346) Nutrition (998) Obstetrics and Gynecology (1144) Occupational and Environmental Health (957) Oncology (3332) Ophthalmology (974) Orthopedics (369) Otolaryngology (420) Pain Medicine (436) Palliative Medicine (130) Pathology (663) Pediatrics (1693) Pharmacology and Therapeutics (691) Primary Care Research (711) Psychiatry and Clinical Psychology (5447) Public and Global Health (9232) Radiology and Imaging (2198) Rehabilitation Medicine and Physical Therapy (1370) Respiratory Medicine (1196) Rheumatology (593) Sexual and Reproductive Health (712) Sports Medicine (530) Surgery (712) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'a00850a23dae4530',t:'MTc3OTU4MzkxOA=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();