DE ananlysis problem of cross-subspecies RNA-Seq
Hi, all
I'm a beginner for transcriptomics. I want to analyze differencial expression of serveral genes between two subspecies (A and B), and have got a RNA-Seq dataset. After mapping the reads to the genome sequences and annotated genes, I encountered some problems about DE analysis:
1. The species have more than 40,000 genes, but less than 20,000 genes can be considered as allelic genes between two subspecies. So if I put the raw read numbers to R package (eg. DEseq, edgeR...) to do DE analysis, more than a half of mapped reads will be missed. Is the results reliable?
2. I also tried to normalize the read numbers with RKPM for each subspecies first. But the RKPM results showed that all alleles in subspecies A have higher expression level than that in subspecies B by pairwised MWU test.
I wonder whether the RKPM normalization is not suitable here or the statistic methods is wrong.
I'm a beginner for transcriptomics. I want to analyze differencial expression of serveral genes between two subspecies (A and B), and have got a RNA-Seq dataset. After mapping the reads to the genome sequences and annotated genes, I encountered some problems about DE analysis:
1. The species have more than 40,000 genes, but less than 20,000 genes can be considered as allelic genes between two subspecies. So if I put the raw read numbers to R package (eg. DEseq, edgeR...) to do DE analysis, more than a half of mapped reads will be missed. Is the results reliable?
2. I also tried to normalize the read numbers with RKPM for each subspecies first. But the RKPM results showed that all alleles in subspecies A have higher expression level than that in subspecies B by pairwised MWU test.
I wonder whether the RKPM normalization is not suitable here or the statistic methods is wrong.
Comment