How to incorporate chimeraslayer in pick_open_ref_otu_picking

[Rittik Deb]

Hi,

I am using pick_open_ref_otu_picking. I want to incorporate chimeraslayer to check for chimeric sequences.

I have obtained non_chimeric_rep_set_aligned_pfiltered.fasta using the protocol suggested by the chimeraslayer page of qiime. However, I have to obtain a biom file for farther downstream analysis. But I am unable to convert the present output file (non_chimeric_rep_set_aligned_pfiltered.fasta) to the biom file needed for farther analysis.

I know that I have to use

make_otu_table.py -i otu_map.txt -t tax_assignments.txt -o otu_table.biom 
to generate the biom file. However the otu_map.txt and tax_assignments.txt are output of pick_otus.py and assign_taxonomy.py 
which are bypassed (or rather collated) when I run pick_open_ref_otu_picking. So how do I generate a biom file without the chimeric
sequences.

Any help will be greatly appreciated.

With warm regards

Rittik

[Rittik Deb]

Hi,

It is probably only me and I apologize for it, but is this is a basic problem which I am unable to solve. Can anyone suggest how everyone check for chimeric sequences using chimeraslayer using qiime.

I am stuck in this step for days now without any outcome. I apologize for any idiotic mistake I am doing but any help will be greatly appreciated

Regards

Rittik

[justink]

Looking into it Rittik, will let you know if I find anything.

[Rittik Deb]

Thanks a lot for your time and concern. Will eagerly wait for any suggestion

[justink]

So, I don't have an authoritative answer, but here's what I suggest:

If you've run pick_open_reference_otus.py, there's likely and output pynast aligned sequences file. That's probably only a small subset of your otus—the ones that weren't reference based. Run identify_chimeric_seqs.py on that file, and any chimeric otus you find there, exclude from your otu table with filter_otus_from_otu_table.py .

Note that this won't check any reference otus. But *hopefully* if a sequence has a high enough identity to a reference sequence, it's not chimeric.

Best of luck.

[Jai Ram Rideout]

Hi Rittik,

You should use the pynast_aligned_seqs/rep_set_aligned.fasta file that's in the output directory created by pick_open_reference_otus.py. For -a, use the same PyNAST template alignment that was used for aligning the sequences. You can find what template alignment was used by looking for "pynast_template_alignment_fp" in the log file in your pick_open_ref output directory. That field should be located near the beginning of the file and will contain an absolute path to the template alignment. Note that the template alignment may be different depending on what version of QIIME you're using (it used to be core_set_aligned.fasta.imputed, now it is 85_otus.pynast.fasta). As long as you use the template alignment indicated in your log file you should be fine.

For example, on my laptop with MacQIIME 1.9.1:

identify_chimeric_seqs.py -m ChimeraSlayer -i pick_open_ref_output/pynast_aligned_seqs/rep_set_aligned.fasta -a /macqiime/anaconda/lib/python2.7/site-packages/qiime_default_reference/gg_13_8_otus/rep_set_aligned/85_otus.pynast.fasta -o chimeric_seqs.txt

Next, filter the alignment to remove chimeric sequences:

filter_fasta.py -f pick_open_ref_output/pynast_aligned_seqs/rep_set_aligned.fasta -o non_chimeric_rep_set_aligned.fasta -s chimeric_seqs.txt -n

Apply lanemask filtering to the non-chimeric alignment:

filter_alignment.py -o non_chimeric_filtered_alignment -i non_chimeric_rep_set_aligned.fasta

Build your phylogenetic tree:

make_phylogeny.py -i non_chimeric_filtered_alignment/non_chimeric_rep_set_aligned_pfiltered.fasta -o non_chimeric_rep_set.tre

Filter chimeric sequences from your OTU table:

filter_otus_from_otu_table.py -i pick_open_ref_output/otu_table_mc2_w_tax_no_pynast_failures.biom -o non_chimeric_otu_table.biom -e chimeric_seqs.txt

Let us know how it goes!

Jai

[Rittik Deb]

I will quickly try out the suggestion and let you know.....