ERROR using CIBERSORTx Docker version

Question

[Leite]

Dear Biostars,

I have been working with the web version and the docker version of CIBERSORTx.

However, I have been getting a very strange result in the docker version, and I don't know what I am doing wrong. Even using the same data and parameters as the web version.

The strange thing is that it results with negative numbers in the docker version.... what can I be doing wrong?

The code I am using in the Docker version:

docker pull cibersortx/fractions

docker run -v D:\src\data:/src/data -v D:\src\data:/src/outdir cibersortx/fractions --username [e-mail] --token d0db6f40f5b632 --mixture My_mixture_1.txt  --sigmatrix My_sigmatrix.txt --perm 100 --absolute TRUE --rmbatchBmode TRUE --QN TRUE  

2023-03-22 15:25:25 >Running CIBERSORTxFractions...
2023-03-22 15:25:25 >[Options] username: [e-mail]
2023-03-22 15:25:25 >[Options] token: d0db6f40f5b632
2023-03-22 15:25:25 >[Options] single_cell: FALSE
2023-03-22 15:25:25 >[Options] sigmatrix: My_sigmatrix.txt
2023-03-22 15:25:25 >[Options] mixture: My_mixture_1.txt
2023-03-22 15:25:25 >[Options] perm: 100
2023-03-22 15:25:25 >[Options] rmbatchBmode: TRUE
2023-03-22 15:25:25 >[Options] QN: TRUE
2023-03-22 15:25:25 >[Options] absolute: TRUE
2023-03-22 15:32:37 Read 1169 items

Thank you in advance for any help!!!

Giuseppe

[Sam]

In the description of the Docker CibersortX parameters which I have, it is said

--QN <bool> Run quantile normalization [default: FALSE]

According to this description at least, if working with RNA-Seq data, one should use --QN FALSE. You can try changing QN parameter, and seeing whether it causes the results to be the same (would appreciate if you would update).