tenglongli / combatcorr Goto Github PK

Dear Mr. Li

I have 1) expression data (rows for genes, columns for samples) and 2) information about samples (Treatment, batch) only in .txt format. And I need to perform the batch-effect correction. Seems to be your batch-adjustment method could be very helpful for my purpose but I cannot understand several moments:

1. Your example is made for two groups(cancer/normal)? How can I change the script for 4 different groups ("C4w","TR4w","TN4w","TH4w")

For this function : adjusted_pValues = design_adj(d,mod,mod0,treat = treat,batch = batch) the treat argument is required, but I doтt know how to get if if there are 4 groups.

2. Can I somehow get the (ComBat+Corr)-adjusted expression data for using it later with limma package?

I will be very grateful to any I advise.

Best regards, Alexandra

I wanted to find the specific GLS method that you applied for statistical inference after getting the clean data but failed.
I am wondering if you utilized lmer or lme. Specifically, which options and which input was put in to fit the model? It would be best if you could give some example code.
I also want to use and cite this method in my research.
Thank you in advance.

Hi Mr.Li,

Your work seems to be a beautiful way for the unbalanced model; however, it seems that the codes cannot provide the adjusted data yet. Could you recommend some ways to use the approximated sample correlation matrix to get batch-free data?

Thank you very much for your kind help in advance.

Ke

Dear Mr. Li,

I'm pretty interested in simulation data generation and want to use this part for my research. But I noticed that no license is added to this repository. So, Can I use code in this repository for my research?

Thanks for your attention.