Variant calling from rnaseq

Question

[Julián]

I am doing a variant calling analysis using rna seq and I feel a bit insecure sometimes, that's why I wanted to ask. Is it a robust analysis?

Also I want to try to draw conclusions by doing a differential expression analysis of genes (using GEO2r) and isoforms (I am using stringTie and ballgown). If you have any opinion about this, I would appreciate it.

Greetings from Argentina.

[ATpoint]

Please ask a precise question. RNA-seq is inferior to DNA-seq for variants, please see many previous threads, but will this stop you from doing it? Why isoform analysis? It's more robust to do gene level.

[Julián]

Ok, thank you for you time.

Answer 1

[james.hawley]

You can, theoretically, identify some mutations from RNA-seq data. But RNA is not DNA, and calling mutations from DNA sequencing data is better for this purpose.

I would recommend checking out GATK's RNA-seq short variant discovery page, as well as the many Biostars threads on this topic:

• Rna-Seq Data Variant Calling
• Variant calling on RNA-Seq data
• Variant calling from RNA-seq?

[Julián]

Yes, I'm doing the analysis following that workflow for this, but as I didn't see it very cited, I sometimes have doubts about its robustness. Thank you very much for your answer.