Hi there,
I’ve trying to play with the minimum length of my reads on the trim.seqs command, but no matter the value I choose, the minimum length I get is always around 250bp, even knowing that the original sff files had a lot of shorter reads. Is the shhh.flows command getting a rid of then?
In addition, why do we choose a minimum of 200bp (as per the SOP) if other platforms like Illumina yield reads of 100bp that can be classified at higher taxonomic levels (Phylum/Class) and still be used for relative abundances comparison?
Thanks a lot,
Marcio.
I’ve trying to play with the minimum length of my reads on the trim.seqs command, but no matter the value I choose, the minimum length I get is always around 250bp, even knowing that the original sff files had a lot of shorter reads. Is the shhh.flows command getting a rid of then?
By default, trim.flows gets rid of sequences with fewer than 450 flows (not the same as bases). You can change this using the minflows and maxflows parameters. However, if you don’t have your own mock community data from the same run, I wouldn’t change anything.
In addition, why do we choose a minimum of 200bp (as per the SOP) if other platforms like Illumina yield reads of 100bp that can be classified at higher taxonomic levels (Phylum/Class) and still be used for relative abundances comparison?
Please see our PLoS ONE paper on the pipeline. For 454, we found that reads shorter than 200 bp had a high error rate.