Entering edit mode
8.5 years ago
gavinmdouglas
▴
10
Hey,
I'm interested in filtering out chimeric reads from 18S sequencing data using UCHIME (reference-based). However, the only eukaryote 18S database to use as a reference I can find is SILVA, which is much larger and not recommended by the UCHIME author. Are there other eukaryote 18S databases available which are of higher quality?
Thanks!
Gavin
Interesting, did the UCHIME author mention why Silva isn't preferred? There aren't a lot of other options, I would expect they would recommend an alternative.
The author discusses it here, apparently it contains "many low-quality sequences which degrade detection accuracy".
I should also say that only 16S and ITS sequences were tested with UCHIME so the author doesn't talk about 18S at all from what I can see. I don't see why the algorithm would be any worse for 18S sequences though.
The impression I get from the QIIME and mothur users/devs is the opposite, that Silva is preferable over the others. See this link.
It may be a matter of filtering to remove the low-quality data; for example the versions of Silva SSU data for mothur are filtered to remove problematic sequences (see under 'Screening the sequences').
Thanks for the info.