Question: What is the consequence of normalizing reads before and during assembly run

[Danielle DeLeo]

I have run various tests to improve my assemblies of tissue-specific transcriptomes. I inadvertently ran trinity (v.2.6.5) with in silico normalization on the species-level assemblies using pre-normalized reads (BBnorm), done discretely for each sample. The assembly stats look better that running just trinity in-silico normalization alone. My question is what are the consequences of normalizing twice? 

For more context, I have 3-4 replicates/samples per species, and found that normalizing the reads prior to trinity with BBnorm improves N50 and BUSCO scores regardless of trinity version I use (current vs old installation). I believe this is because these are crustaceans and various studies have shown K=23 is optimal for this group. 

Appreciate any insight or advice before moving forward, thanks! 

[Brian Haas]

Hi Danielle,

It could be that bbnorm does a better job at normalization than what Trinity does. If bbnorm is eliminating poor quality reads or doing error correction, than that could explain it.  The second normalization done by Trinity might not be doing much unless the bbnorm coverage cap is a lot higher than what Trinity uses (200x in the current Trinity).  You could always disable the Trinity normalization via --no_normalize_reads if the bbnorm alone is sufficient.... but if it's not, then it could lead to runtime performance problems w/  Trinity.

--
You received this message because you are subscribed to the Google Groups "trinityrnaseq-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to trinityrnaseq-u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/trinityrnaseq-users/bfbeddc3-4f6b-4943-a5f7-775c0724c973n%40googlegroups.com.

--

--
Brian J. Haas
The Broad Institute
http://broadinstitute.org/~bhaas