I would like to ask why in RNA seq analysis (alignment step) we use sometimes reference genome instead of reference transcriptome? thank you!
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
11
Pseudoaligmnent to a reference transcriptome with something like Kallisto is perfectly valid, assuming you are in an organism whose reference transcriptome is very well documented.
But if you want to find novel transcripts or novel splicing junctions, you need to align to genome.
-
$\begingroup$ thank you! Could I use STAR or a similar program (not Kallisto or Salmon etc) with reference transcriptome? $\endgroup$– mariluCommented Aug 5, 2020 at 20:57
-
2$\begingroup$ You can make a STAR reference out of a transcriptome, but why would you do that? It wasn't designed to be run like that. $\endgroup$ Commented Aug 5, 2020 at 21:07
-
$\begingroup$ if i am only interested in differential expression, could i use star with reference transcriptome to decrease computational effort? $\endgroup$– mariluCommented Aug 5, 2020 at 21:24
-
2$\begingroup$ As a rule, do not use programs for things they were not designed for unless you are absolutely sure you understand what you are doing and why. If you want something faster than STAR, use Kallisto. $\endgroup$ Commented Aug 5, 2020 at 21:34
-
$\begingroup$ I actually think you could run STAR on a transcriptome. In their manuscript, you can run alignment using a GTF file. $\endgroup$ Commented Aug 5, 2020 at 21:36