Gene expression data and clinical information of CRC cohorts used in this study were acquired from UCSC Xena (15) and Gene Expression Omnibus (GEO) (16). Two CRC single-cell datasets were downloaded from GEO under the accession number GSE178341 and GSE108989. The metastatic urothelial cancer treated with anti-PD-L1 agent (atezolizumab) cohort IMvigor210 was obtained via R package IMvigor210CoreBiologies (17). The melanoma cohort undergoing anti-PD-1 checkpoint inhibition therapy was downloaded from GEO under accession number GSE78220. Only primary tumors and treatment-naive patients were included.
Microarray data (GSE103479, GSE17538, GSE33113, GSE37892, GSE38832, GSE39084 and GSE39582 used to develop IRScore; GSE39395 and GSE39396 used to draw signatures of fibroblasts, endothelial cells and epithelial cells) were downloaded and processed using the R package GEOquery (18). For TCGA cohort, log-transformed (on a base 2 scale) gene expression data were downloaded from UCSC Xena (https://xenabrowser.net/datapages/). Processed expression data of GSE178341, GSE108989 and GSE78220 were directly downloaded from GEO. Counts data and patients’ information from the IMvigor210 cohort were obtained by function counts and pData, and the gene counts were transformed into TPM for the following analysis. All datasets used in this study were listed in Table S1 .
Do you have any questions about this protocol?
Post your question to gather feedback from the community. We will also invite the authors of this article to respond.