hCINAP expression in colorectal cancer

The authors declare that the data analysis processes
supporting the findings of this study are available within the article and its
Supplementary Information files. The TCGA gene expression profile data, as
recomputed based on gencode v23, were downloaded from UCSC Xena (http://xena.ucsc.edu/). The TCGA clinical data were downloaded
from the GDC Data Portal (https://gdc-portal.nci.nih.gov/), with accession number phs000178.v9.p8 in
dbGap.

Supplementary Information:

For analyzing the hCINAP
expression in CRC, we downloaded the recomputed TCGA gene expression datasets for
COAD and READ cancer types from the UCSC Xena (http://xena.ucsc.edu/). The gene model was based on gencode v23,
and the expression unit is TPM (Transcript per million). The clinical data were
downloaded from the GDC Data Portal (https://gdc-portal.nci.nih.gov/).

For differential expression analysis, we compiled a selected
sample set, including 367 tumor- and 51 normal-samples, in which each sample
has information available for clinical variables such as gender, age and race (Supplementary
Table1). For expression analysis by pathological stages, we only used those
tumor samples with stage information (Supplementary Table1). The dataset used
for profiling gene expression by CRC subtypes was compiled based on the results
of consensus molecular subtypes (CMSs) described previously [PMID: 26457759] ,
containing 265 tumor samples (Supplementary Table1).