Open wang99999shang opened 2 years ago
Hello,
You would want to use the returnSigMatrix() function for this. Its default arguments are as described below.
Previous versions wrote the signature matrix to a file during a standard deconvolution run, but it looks like this was unintentionally removed in 2.0. I will have to add it back when I get around to updating to 2.1
returnSigMatrix( InputMatrix, #Full expression matrix from which a signature matrix will be created minPerCT = 50, #Take no fewer than 50 signature genes for any cell type TotalSigs = 50, #Take an average of 50 signature genes per cell type strModes = "Entropy" #select signature genes by minimum entropy)
I have now added this functionality to GEDIT2.1. The program will now write the signature matrix by default, alongside wherever the results are saved as determined by the -out parameter. To also write the "scaled" mixture and reference files (the exact matrices that are given to the linear regression), indicate -outFile All. To write only the predictions, and omit the signatures matrix, indicate -outFile None
Hello! Thank you for building the GEDIT program. Can you give an example for how to use scripts to generate signaturematrix? I don't know whether I have got the right way to use getSigGenesModal.py