CIBERSORT input file for RNA-Seq data

Question

[MS]

Hi,

I am trying to figure out CIBERSORT for RNA-Seq data. I have 5 control samples and 32 patient samples. I have DeSeq2 outputs and count data.

I have 2 questions.

1.Should I use DeSeq2 outputs as input to CIBERSORT for Impute Cell Fractions mode or just should I use count data obtained from Featurecounts?

2.Should I include control group data into input file or just patient's data

[DareDevil]

I recommend using tpm matrix for CIBERSORT algorithm

[Sara]

Hi,

I am also trying to use CIBERSORT for RNA-Seq data. I put my RNA-Seq count file as an input in the Mixture file.

I have some problems with the signature matrix, and class label files (phenotype classes file) how do I have to generate them? (it would be for the brain)

Thank you in advance!

Answer 1

[William]

Hi,

First question: You should use the normalized matrix with the values for all transcripts from DESeq2 as CIBERSORT input. You will need to extract it using the #count() function and use the gene symbols in the first column. Write it in your current directory and upload in CIBERSORT.

Second question: You should use all data in the CIBERSORT, because it will estimate the cellular abundance for each variable (Patient/column). After the analysis, you can plot it and split the studies groups as you want.

all the best,

[MS]

Thank you so much for your help!