Checking workflow for chimera slayer

[Riley Hughes]

Hello, I am trying to figure out if I have the correct workflow for getting a biom file out of QIIME1 using chimera slayer as the chimera checking step. I typically use a different workflow so this is new to me. My workflow is attached but basically I pick_open_reference_otus.py > identify_chimeric_seqs.py (using rep_set_aligned.fasta) > filter_fasta.py (using rep_set and non_chimeric_rep_set_aligned.fasta and chimeric_seqs.txt) > make_otu_table.py (using final_otu_map_mc2.txt and chimeric_seqs.txt).

Any advice would be greatly appreciated! Thank you

[TonyWalters]

Hello Riley,

That does look correct (and matches the guide here http://qiime.org/tutorials/chimera_checking.html#chimeraslayer).

-Tony

[Riley Hughes]

Great, thank you! Just wanted to double check.

[Riley Hughes]

Hello all, just following up on this topic. I have been trying to run this but the chimeras.txt file is showing up blank. Does anyone know why this might be happening? I use the same fasta reference file for both pick OTUs and identify chimeras (97_otus.fasta). I also don't think it is a memory issue because I ran it on a interactively on a computer with 512GB of RAM with no other competing processes. The process ran to completion and did not give any errors. Any help is appreciated! Thank you.

[TonyWalters]

Hello Riley,

There's a thread here (a bit lengthy, but there is a "positive control" test file in there to see if ChimeraSlayer will complete and detect chimeras):

https://groups.google.com/forum/m/#!msg/qiime-forum/6RNK_V4I3Dk/5HFovx9mAgAJ

It may be worth testing that just to confirm that it is completing-there may be the possibility that it's not detecting chimeras too. If this is the case, you might also try the usearch61 (alternatively vsearch) approach to see if chimeras are detected with that software instead.

-Tonu

[Riley Hughes]

Hi Tony, thank you for your help. I did try the positive control and it worked but my data still produced a blank chimeras.txt file. I've been scouring the internet but can't find anything that might help. Would it be possible to send you my rep_set_aligned.fasta file to see if it works with someone else's installation? Otherwise, do you have any other advice or solutions I could try? Thank you very much, I apologize for the inconvenience!

[TonyWalters]

Hello Riley,

Unfortunately there really hasn't been any updates for that software for a long time. You could try the usearch61 (alternatively vsearch as a plugin) approach as another method, as it's a bit newer.

Another option would be one of the ways that was done a long time ago-filter out rare OTUs (normally PCR-based chimeras would be low frequency), and for any OTU that of interest, e.g., one that's significantly different in abundance, take the representative sequence from the rep_set.fna file and blast different fragments of it (e.g. 3 different 100 base pairs of it) and confirm that the different fragments all hit the same taxa. Sometimes there are multiple taxa hit at high % identity, and that's fine, as long as two of the fragments aren't unambiguously hitting different taxa.

[Riley Hughes]

Ok thank you so much for your help, I greatly appreciate the input!