Robustness of regulatory colocalization evidence at major depression risk loci — corrective release v2

Abstract

CORRECTION OF VERSION 1 This is a corrective release, not an extension of the original findings. The earlier colocalization results are superseded because inappropriate variant-count normalization in posterior combination and cross-gene QTL record selection affected the analyses. The former 130-locus “shared mechanism” category was an operational classification, not evidence of established shared biological mechanisms. Earlier numerical summaries, including 130, 56, and 50.0%, and their affected downstream interpretations must not be reused as validated findings. Corrected gene-specific analysis yields 10,966 gene–locus–tissue tests, 105 default-positive tests involving 88 genes at 55 loci, four tests at three loci stable under all six prior configurations, and cross-resource concordance of 34/81 matched primary-positive tests. These denominators refer to tests rather than independent loci. This release accompanies “Robustness of regulatory colocalization evidence at major depression risk loci.” It contains corrected analysis results, source code, figures, manuscript and supplemental materials, provenance records, numerical validation, and aligned inputs for 417 multi-signal candidates. Version 1 is retained for transparency and the audit trail. The results support a restricted set of regulatory associations. They do not establish tissue-shared mechanisms, verified independent replication, disease specificity, or therapeutic efficacy. Secondary GTEx v10 and immune analyses cover selected subsets. Exact person-level overlap between source studies is not quantified. Missing evidence and non-assessable multi-signal results are not treated as negative findings. Project-authored text and derived results are distributed under CC BY 4.0. Software retains the historical MIT license. Third-party materials remain subject to their providers’ terms; see the included LICENSE and provenance documentation. This work was supported by the General Project of the Zhejiang Provincial Department of Education, China (grant Y202559996), and the Basic Scientific Research Project of the Wenzhou Science and Technology Bureau, China (grant YC20250576). The authors declare no competing interests. Correspondence: Cheng Zhu, School of Mental Health, Wenzhou Medical University, Wenzhou, China.Email: wmuzc@wmu.edu.cnAlternative email: zhucheng871005@163.comORCID: https://orcid.org/0009-0009-2146-5759 All versions: https://doi.org/10.5281/zenodo.22858380Historical version 1: https://doi.org/10.5281/zenodo.22858381