I am analyzing a scRNAseq data trying to follow a protocol suggested in this paper- https://www.nature.com/articles/nature24489 mentioned under Computational analysis The statement reads- " Selection of variable genes was performed by fitting a generalized linear model to the relationship between the squared coefficient of variation and the mean expression level in logarithmic space, and selecting genes that deviated significantly (P < 0.05) from the fitted curve"
I have two naive questions: 1. In general GLM can be done in two data columns. However in my case there are several columns of data corresponding to single cells. How can I use the matrix for GLM. I looked into JMP and past3 documentation.
- What is the best GUI based software or an R package that can fit the model for multiple columns data and out put p value for each row (gene).
Apologies, if it looks like a very naive question. However, I tried but could not overcome this bump.