A transcriptome-wide association study integrating multi-omics bioinformatics and Mendelian randomization reveals the...

An abundant amount of colon cancers is diagnosed every year, accounting for 9% of malignant tumors. Even with the progress of relevant research, the 5-year survival rate for colon cancer is still less than 60%, indicating that improving the prognosis of colon cancer is still a challenge that needs to be overcome. This study employed the algorithm "scissor" to integrate the single-cell sequencing data and bulk transcriptome data with prognosis information to predict prognosis-associated cells (PAC). Summary-data-based Mendelian randomization (SMR) analysis was conducted using expression quantitative trait loci data and GWAS data to identify genes having causal associations with prognosis phenotype in colon cancer patients and five traditional two-sample Mendelian randomization methods were utilized to confirm the results. Finally, our findings were validated based on two independent external validation datasets, GSE17536 and GSE39582. The real-world tissue dataset with corresponding immunohistochemical (IHC) experiments was utilized to confirm our findings. We determined that the majority of PACs were fibroblasts. On top of that, this study identified ADAMDEC1 as a gene that has a significant causal association with overall survival. ADAMDEC1, highly expressed in highly differentiated fibroblasts, was ascertained its high expression was linked with a better prognosis of patients with colon cancer by the related bulk transcriptome analysis. Our dataset presented that higher IHC scores were associated with a better prognosis for colon cancer, further validating our results. This study has identified ADAMDEC1 as a prognostic protective factor for patients with colon cancer, providing clues for clinical trials and drug experimental target research.