This method is created as described in the manuscript - "Deconvolution of bulk gene expression profiles from complex tissues to quantify subsets of immune cells".
The tool is designed to estimate the fractions of immune cells from the expression profiles of tissues, such as tumor masses, PBMCs, etc. The 8 immune cells treated in this method include naïve CD8 T cells, naïve CD4 T cells, alternatively activated macrophages, classically activated macrophages, regulatory T cells, T helper cells, natural killer cells, and dendritic cells.
Fixed. Now redundant records for each gene will be removed, and only the record with the higher value will be taken for the cell composition deconvolution. (05/02/2020)
- R
- python 3
The data provided in Pre.reference.csv is used as the reference gene expression signatures (RefGES) for the immune cell types to be treated in this method.
Before starting the cell deconvolution analysis, the input expression profiles with initially unknown cell composition must be pre-processed by running pre-process.R, which could retrieve the expression levels of the common genes shared by the input data and our RefGES. Two files, namely Sample.csv and Reference.csv, will be automatically generated by pre-process.R. Then it is ready for us to run SCORE.py, which will take the above 2 files as the input files to infer the cell composition in the our samples.
After running SCORE.py, three output files will be generated, including score.csv, bar.pdf and boxplot.dpf. score.csv provides the inferred fractions of the 8 types of immune cells. bar.pdf and boxplot.pdf provide the visualization of the inferred cell fractions.
In order to allow for a quick trial of this method by the users, we provide an example file, example_expression_profile.csv, which can be used as the input file to infer the cell composition in the sample.
Rscript pre-process.R -i "file name of our raw data" # you may try our `example_expression_profile.csv`
python SCORE.py
Yen-Jung Chiu (holiday01) and Yen-Hua Huang created this method. holiday01 implemented the code.
Chiu, Y., Hsieh, Y. & Huang, Y. Improved cell composition deconvolution method of bulk gene expression profiles to quantify subsets of immune cells. BMC Med Genomics 12, 169 (2019) doi:10.1186/s12920-019-0613-5