Read correction prior to assembly

[kmbultman]

I am working with on both Illumina-only and hybrid assemblies and plan to use Unicycler to perform the de novo assemblies. Do you recommend using a read correction program such as Trimmomatic to do quality control on the reads prior to assembly or is the read correction through SPAdes or Racon sufficient? I plan to use these genomes for comparative genomic analysis downstream and want to have the best assemblies possible.

Thanks!

[dswan]

I've seen 'trimmomatic' being described as read-correction in more than one post recently. This is not what trimmomatic does, it just trims reads, removes adapters, and applies quality thresholds to reads. Yes, this should be run prior to assembly, because the actual read correction done through SPAdes (BayesHammer) is very different to what trimmomatic does.