Ulcerative colitis (UC) is a prevalent chronic gastrointestinal disease. Gene plays an important role in UC pathogenesis. Therefore, we aim to identify UC susceptibility genes and specific cell types expressing these genes. We conducted a cross-tissue transcriptome-wide association study by integrating UC GWAS with 49 tissues gene expression matrix from the Genotype-Tissue Expression project. Subsequently, we employed Functional Summary-based Imputation to verify candidate genes within colon tissue. Conditional and Joint Analysis was utilized to filter out genes potentially influenced by linkage disequilibrium. Multimarker Analysis of Genomic Annotation was then applied to pinpoint genes relevant to UC. Validation of the selected genes was performed using Mendelian randomization. GeneMANIA analysis was conducted to elucidate biological functions of identified genes. Finally, single-cell RNA sequencing was employed to ascertain cell types in which these genes are enriched. The cross-tissue transcriptome-wide association study, Functional Summary-based Imputation and Multimarker Analysis of Genomic Annotation analyses identified a total of 5 genes, of which 3 genes, ADCY3, ITGB6, and MTMR3, were retained after Mendelian randomization. These genes were found to be implicated in several functional pathways, including the cyclic AMP metabolic process and phosphorus-oxygen lyase activity. Furthermore, we observed ADCY3 predominantly enriched in B cells, while ITGB6 and MTMR3 enriched in epithelial cells. Our study has identified 3 genes associated with UC susceptibility. These findings not only enhance our understanding of the genetic underpinnings of UC, but also offer novel avenues for exploring molecular mechanisms and potential therapeutic targets for UC.

The change of plasma lipid species has close contacts with gastric cancer (GC). However, the specific mechanism still needs to be explored further. We aim to utilize plasma proteins to decipher the association between lipid species and GC, and seek possible drug targets for GC. We performed a two-step Mendelian randomization (MR) analysis to investigate causal relationships among 179 lipid species, 4907 plasma proteins, and GC. Using summary-data-based MR and colocalization, we first examined protein–GC associations in discovery (N = 35,559) and validation (N = 54,219) cohorts. Subsequent MR analyses assessed lipid–GC and lipid–protein relationships, followed by mediation analysis using error propagation methods. Finally, macromolecular docking of prioritized proteins identified potential therapeutic ligands. Our MR analysis revealed causal relationships between 12 lipid species and GC, as well as 3 plasma proteins and GC. Importantly, mediation analysis demonstrated that CCDC80 protein mediates 2.90% (95% CI: 0.30–5.5%) of the protective effect of diacylglycerol (16:1_18:1) against GC. Based on these findings, we identified valproic acid as a promising therapeutic candidate targeting CCDC80 for GC treatment. Our study demonstrates that reduced CCDC80 expression mediates the tumor-promoting effects of diacylglycerol (16:1_18:1) in GC pathogenesis. Molecular docking confirms valproic acid binds stably to CCDC80, suggesting its therapeutic potential. These findings advance GC etiology understanding and provide a new drug development direction.