Identification and verification of prognostic cancer subtype based on multi-omics analysis for kidney renal papillary cell carcinoma.

BACKGROUND

Identifying Kidney Renal Papillary Cell Carcinoma (KIRP) patients with high-risk, guiding individualized diagnosis and treatment of patients, and identifying effective prognostic targets are urgent problems to be solved in current research on KIRP.

METHODS

In this study, data of multi omics for patients with KIRP were collected from TCGA database, including mRNAs, lncRNAs, miRNAs, data of methylation, and data of gene mutations. Data of multi-omics related to prognosis of patients with KIRP were selected for each omics level. Further, multi omics data related to prognosis were integrated into cluster analysis based on ten clustering algorithms using MOVICS package. The multi omics-based cancer subtype (MOCS) were compared on biological characteristics, immune microenvironmental cell abundance, immune checkpoint, genomic mutation, drug sensitivity using R packages, including GSVA, clusterProfiler, TIMER, CIBERSORT, CIBERSORT-ABS, quanTIseq, MCPcounter, xCell, EPIC, GISTIC, and pRRophetic algorithms.

RESULTS

The top ten OS-related factors for KIRP patients were annotated. Patients with KIRP were divided into MOCS1, MOCS2, and MOCS3. Patients in the MOCS3 subtype were observed with shorter overall survival time than patients in the MOCS1 and MOCS2 subtypes. MOCS1 was negatively correlated with immune-related pathways, and we found global dysfunction of cancer-related pathways among the three MOCS subtypes. We evaluated the activity profiles of regulons among the three MOCSs. Most of the metabolism-related pathways were activated in MOCS2. Several immune microenvironmental cells were highly infiltrated in specific MOCS subtype. MOCS3 showed a significantly lower tumor mutation burden. The CNV occurrence frequency was higher in MOCS1. As for treatment, we found that these MOCSs were sensitive to different drugs and treatments. We also analyzed single-cell data for KIRP.

CONCLUSION

Based on a variety of algorithms, this study determined the risk classifier based on multi-omics data, which could guide the risk stratification and medication selection of patients with KIRP.