Revisit linear regression-based deconvolution methods for tumor gene expression data

Table 1 Associations of CIBERSORT estimates for closely related features

	Naive versus memory B cell (expected r = 0.7)	Activated versus resting CD4 memory T cell (expected r = 0.3)
Glioblastoma	r = −0.34 ⍴ = −0.77	r = −0.05 ⍴ = −0.06
Kidney renal clear cell carcinoma	r = −0.07 ⍴ = −0.29	r = −0.12 ⍴ = −0.13
Lung squamous carcinoma	r = 0.13 ⍴ = −0.38	r = −0.26 ⍴ = −0.19
Lung adenocarcinoma	r = −0.07 ⍴ = −0.37	r = −0.29 ⍴ = −0.26

The expected correlations were obtained from Fig. 1a and are given in the column headings. RNA-seq data of each cancer type were analyzed by the CIBERSORT online server to predict fractions of immune subsets. Pearson’s r and Spearman’s ρ were calculated between two pairs of closely related features: (1) naïve and memory B cells, and (2) activated and resting memory CD4 T cells. Red values indicate a negative association

ISSN: 1474-760X