Re: incomplete biom file
44 views
Skip to first unread message
Message has been deleted
Falk Hildebrand
Apr 28, 2016, 5:48:33 AM4/28/16
to LotuS rRNA pipeline
Hey Jan,
can you please first check that the OTU matrices (OTU.txt) between the two runs contain also the same number of OTUs (6120) and post the two lotusrun logfiles from both runs? The new version has more quality filters, but I also want to make sure that the biom file was correctly created (i.e. the OTU.txt should correspond to the OTUs in the biom file).
cheers,
Falk
Jan Söderman
Apr 28, 2016, 6:48:23 AM4/28/16
to LotuS rRNA pipeline
Hi Falk,
I can confirm that the OTU.txt files between the runs do contain the same number of OUTs.
I have attached the requested files. You will notice that the first Lotus run (old) was ran using the -refDB, -i, -m, -o, and -s options. I later realized that the output indicate "Sequencing platform=454", even though I have used the MiSeq-version of the sdm-file and I have paired-end data. It was because of this I wanted to reanalyze my files, however, this time using also the -p and the -thr options. I have checked that the discrepanies between the old and the new Lotus versions are not due to the use of these additional options. Analysing the data with or without the -p and -thr options give the same results using Lotus 1.504.
Best regards,
Jan
I can confirm that the OTU.txt files between the runs do contain the same number of OUTs.
I have attached the requested files. You will notice that the first Lotus run (old) was ran using the -refDB, -i, -m, -o, and -s options. I later realized that the output indicate "Sequencing platform=454", even though I have used the MiSeq-version of the sdm-file and I have paired-end data. It was because of this I wanted to reanalyze my files, however, this time using also the -p and the -thr options. I have checked that the discrepanies between the old and the new Lotus versions are not due to the use of these additional options. Analysing the data with or without the -p and -thr options give the same results using Lotus 1.504.
Best regards,
Jan
Falk Hildebrand
Apr 28, 2016, 10:01:51 AM4/28/16
to LotuS rRNA pipeline
Hey Jan,
this was indeed a bug. I have uploaded version 1.506 to the server, can you please upgrade and just quickly post if this fixes the problem for you? Note that this bug occured since version 1.50 and only if RDP taxonomy was used.
best,
Falk
Jan Söderman
Apr 29, 2016, 3:34:15 AM4/29/16
to LotuS rRNA pipeline
Hi Falk,
I have now updated Lotus to 1.506 using the autoInstall.pl script, and re-ran the data using the following command: ./lotus.pl -refDB SLV -thr 20 -p miSeq -i /path_to_input/ -m /path_to/map_psokal_data_Unix.txt -o /path_to_output/ -s /path_to/sdm_miSeq.txt
As previously (using v1.44 and v1.504) the OTU.txt file contain 6120 OTUs. However, the biom-file still has the wrong shape ("shape": [4366,72], while v1.44 gave the expected "shape": [6120,72]).
I have attached the LotuS_runlog.txt, and, if needed, I can e-mail you the biom-file.
Sincerely,
Jan
Jan Söderman
May 2, 2016, 8:10:05 AM5/2/16
to LotuS rRNA pipeline
Hi Falk,
I can now confirm that the latest lotus.pl version does work. I get a biom-file with the correct shape, and the significant results after phyloseq and DESeq2 are back.
Thank you,
Jan
Reply all
Reply to author
Forward
0 new messages