Jackknife Model Averaging Prediction Methods for Complex Phenotypes with Gene Expression Levels by Integrating External Pathway Information

Table 1

Sample sizes and the number of genes for each cancer in the TCGA dataset used in our analysis.

Phenotypes	Initial gene expression data		Initial clinical data (N)	Final data after quality control
	N	G		N	G
BRCA	1,218	20,531	1,247	1,083	17,675
COAD	329	20,531	551	275	17,493
CRC	434	20,531	736	367	17,510
PAAD	183	20,531	196	178	17,675

Note. N is the sample size and G denotes the number of genes. The average number of genes incorporated in each pathway for the seven phenotypes was 65 (ranging from 1 to 1,139), and about 21% genes belonged to multiple pathways. BRCA: breast cancer; CRC: colon and rectal cancer; COAD: colon cancer; PAAD: pancreatic cancer.