I am Yogender gowtham and I am a graduate student from Clemson University (USA). I am working on understanding the transcriptome of Chinese hamster ovary (CHO) cells using RNA-seq. To perform differential gene expression analysis between my treatments, I am using edgeR software package.
Since I have multiple factors in my treatments, I am using GLM to perform differential gene expression analysis. However I have few questions related to the likelihood ratio test and the likelihood ratio in the output.
1) From what I understand, the likelihood ratio tests compares two models - a full model and a reduced model to see which one fits better. Could someone please tell me what would be the null and alternate hypothesis for this test?
2) Below is a typical output from my pairwise comparison.
It looks like greater the likelihood ratio, lower is the p-value. Is the LR value a estimate from likelihood ratio test? What exactly is it and how does it relate to p-value?