DiseaseCIO: a multi-omics atlas for characterizing clinical intervention outcomes across diverse diseases

Abstract

Heterogeneity in clinical intervention outcomes across patients is a major barrier to precision medicine. The integration of multidimensional high-throughput omics can help reveal the molecular basis of differences in clinical interventions and provide insights into outcome prediction, prognostic stratification, regulatory patterns, and resistance mechanisms. We present DiseaseCIO (Disease Clinical Intervention Outcomes, http://db.edbc.org/DiseaseCIO/), a database that integrates microbiome, transcriptome, and epigenome data with human disease intervention outcomes. It is designed to characterize the molecular landscape associated with disease intervention outcomes at the individual patient level. DiseaseCIO encompasses 42 798 pretreatment samples from 94 disease states, covering 510 treatment regimens involving 377 pharmacological and non-pharmacological interventions. These samples comprise four types of specimens collected from eight sampling locations. Using the customized iMOD workflow for multi-omics data processing, we generated 20 categories of molecular features and identified >1.84 million candidate feature–outcome associations. DiseaseCIO also provides interactive tools for differential analysis, prognostic evaluation, cross-omics correlation analysis, and cross-cohort integrative analysis of multidimensional features as well as enrichment analysis and signature construction for user-submitted feature sets. DiseaseCIO provides a molecular-level resource for exploring intervention outcome heterogeneity across diseases and identifying candidate multi-omics biomarkers associated with treatment response and prognosis.