COVID-19と結腸直腸がんにおける共有分子マーカーとメカニズムの解明：バイオインフォマティクスと機械学習を用いて

Colorectal cancer (CRC) and Coronavirus Disease 2019 (COVID-19) are distinct diseases that may share overlapping molecular mechanisms, particularly in immune dysregulation. However, the specific regulatory pathways driving this shared pathophysiology have remained elusive, as prior studies have been limited by single-level data. To dissect this common pathobiology, we implemented a synergistic computational framework, integrating bulk transcriptomics with single-cell data. Through a multi-tiered analysis pipeline employing differential expression, weighted gene co-expression networks, and machine learning-based feature selection, we pinpointed a core molecular signature of 31 shared hub genes. Among these, four core candidates-GPR15, PTGDR2, FCER1A, and MAL-were significantly downregulated, a finding robustly associated with impaired CD8+ T cell infiltration. Delving deeper into the regulatory architecture using a modified weighted out-degree centrality algorithm, we constructed an integrated transcription factor-microRNA-target network. Network analysis revealed upregulation of p53 and downregulation of miR-3619-5p as possible drivers of immune dysfunction. Finally, E-4031 was identified through molecular simulation as a potential therapeutic agent targeting all four core genes. These findings uncover a shared regulatory axis involving immune suppression and transcriptional disruption, and provide promising diagnostic and therapeutic targets for CRC and COVID-19.