Potential key genes are expected to become biomarker for early diagnosis of colorectal cancer through bioinformatics analysis.

Colorectal cancer (CRC) is the leading cause of cancer-related deaths in the world. The aim of this study was to identify the potential key genes, and associated pathways for early-onset CRC through bioinformatics methods. We integrated the gene expression patterns of CRC from three RNAseq datasets (GSE8671, GSE20916, GSE39582) from GEO database to identify DEGs between CRC and normal samples. We established a gene co-expression network through WGCNA. Through the WGCNA calculation, the genes were divided into six modules. The WGCNA analysis screened 242 genes associated with pathological stage and colorectal adenocarcinoma, 31 of which had the ability to predict OS with an AUC >0.7. The GSE39582 dataset identified 2040 DEGs between the CRC and normal samples. The two were intersected to obtain two genes: NPM1 and PANK3. The two genes were used as a threshold to divide the samples into high group and low group for survival analysis. Survival analysis showed that increased expression of both genes was significantly associated with a poorer prognosis. These two genes (NPM1 and PANK3) could be possible marker genes for early diagnosis of CRC, providing ideas for other experimental studies in the future.