Reconstituting Specific Cell-type Expression
Written by Joseph Ng, October 2018
RESPECTEx takes as input:
- tumour bulk gene expression data,
- estimates of tumour purity of the tumour samples(obtained from e.g. tumour purity estimation algorithms, or IHC-based pathological assessments), and
- estimates of immune cell populations in the tumour samples (obtained from e.g. the CIBERSORT algorithm)
and infer for the given cohort, the mean gene expression levels for specific cell types.
This code is a pipeline which integrates the three input, and performs a non-negative least-square regression taking the bulk expression data as the response and the cell-type proportion estimates as covariates. The mean expression levels per cell type are the coefficients from the regression model.
RESPECTEx.R
: The main RESPECTEx pipeline.GSE75688_ESTIMATEpurity.txt
: Tumour purity estimates for the GSE75688 breast cancer cohort generated using the ESTIMATE algorithm, extrapolated by comparing to TCGA tumours.GSE75688_pooledOrTumor_TPM_normalized_CIBERSORT.txt
: Immune cell subpopulation proportion estimates for the GSE75688 breast cancer cohort generated using the CIBERSORT algorithm.GSE75688_pooledOrTumor_TPM_normalized.txt
: Transcript per million (TPM) expression data from pooled cells/bulk tumour samples of the GSE75688 breast cancer cohort. This is a test input to the RESPECTEx pipeline.GSE75688_RESPECTEx.txt
: Expected output from RESPECTEx analysis of the GSE75688 breast cancer cohort.
Please see RESPECTEx.R
for instructions on how to run the analysis.
RESPECTEx is distributed and licensed under GNU General Public License v3.0. Please see file "LICENSE" for details.
Ng JCF, Quist J, Grigoriadis A, Malim MH & Fraternali F. Pan-cancer transcriptomic analysis dissects immune and proliferative functions of APOBEC3 cytidine deaminases. Nucleic Acids Research, gky1316, 2019.