I am doing an analysis with transcriptome data and comparing the same with the Genomic data. More precisely, I am looking at mRNA results of a tumor sample without having the normal tumor samples mRNA levels to compare with. So the doubts are following:
1. Does a transcript that is found to be mutated (SNPs / Insertions / InDels etc.) in the NGS data is always supposed to show any additional changes at transcriptome level?
2. Other than CNVs, can mutations be verified at the mRNA expression level?
VAF of RNA and DNA will give you a strong correlation. However, you can aspect some changes due to various intermediate processes such as NMD (Non-sense mediated decay), which eliminates the mutations in stop codons, so reduces gene expression error. CNA analysis based on RNA-seq may not be that accurate as much as based on DNAseq. It is better to explore important findings through visualization tools such as IGV. for further details you can refer to https://www.nature.com/articles/s41591-020-1072-4
Amarinder Singh Thind is right, IGV is a great tool to see whats going on in both RNAseq and whole genome sequencing data. To estimate allele frequency in RNAseq, you need to have a fairly large coverage.
Dear Kunal Ghosh Roy ,
How are you?
Here are my thoughts about your questions.
1) Not necessarily. You might have SNPs that will not generate a significant change at the mRNA level. If you compare all possible RNA changes (including ncRNA) you might see. Nevertheless, the SNP change still might not generate something relevant in the end. Generally, SNPs should always be taken lighly. Still, it is always crucial to check because, as mentioned, having both genomic and transcriptomic data is always a good correlation. Insertions and deletions will usually lead to some change at the transcriptional level.
Nevertheless, you mentioned that you have no normal tumoral samples. I never worked with datasets that lack normal samples. Usually, I don't consider "safe" to analyze RNA-seq tumoral data without normal samples. I know that will always depend of what is your goal at the end (such as comparing DEG changes between two tumoral types) but still. If you have a way to find and add normal samples to your study, I would strongly advise you to do it.
Cheers
- Badges
- Science method