I am running the LotuS2 function on usegalaxy.eu using 62 fasta files, a txt file for taxonomy reference lineages and fasta taxonomy reference sequence. I checked the error and the bug icon showed “Fatal error: Exit code 9 ()”. When I emailed about the problem the response was: “We currently recommend to use a conda environment instead of a biocontainer to prevent this problem.” I am wondering what steps should be taken to fix this?
Hi @smoss33
Let’s see if we can help here! I am not sure who you asked – the developers themselves? I wonder if the implementation in Galaxy is a bit different (on purpose) or if there is a something else we can tune to avoid that error.
I just ran a new test in this history with both of the tool tests for the Galaxy wrapper. Can you compare how you set up your run, and explain what you did different? So I can replicate the error and get it reported (if needed) or help you to solve your error? I used the most current version of the tool since older versions sometimes had problems (leading to the updated versions!).
- LotuS2 fast OTU processing pipeline (Galaxy Version 2.32+galaxy0)
- earlham-galaxytools/tools/lotus2 at main · TGAC/earlham-galaxytools · GitHub
lotus2.xml
is the tool wrapper, and at the end includes the tool teststest-date
is where to find the files for the tool test
- My version of those tests run https://usegalaxy.eu/u/jenj/h/test-lotus2
You can also share your history back if you want (how we usually start troubleshooting). You also might want to try to subset your data, and run a smaller test with your parameters and reference data.
I don’t know if the container environment can be adjusted or not, but the EU server is undergoing maintenance in a few days. We can follow up once all this is confirmed.
Let us what you think!
Hi @smoss33,
I found your bug reports in our (Galaxy Europe’s system) and took a look at your failed runs. With that info, not available to @jennaj, I think it’s clear what’s happening.
The tool produces a main log file, which can be viewed even for failed jobs and in your case it has the section:
SDM options /usr/local/tools/_conda/envs/__lotus2@2.32/share/lotus2-2.32-0/configs/sdm_miSeq.txt
TempDir tmp_folder
NumCores 4
------------ Pipeline config --------------
Sequencing platform miseq
Amplicon target fungi, SSU
Dereplication filter -derepMin 10:1,5:2,3:3
Clustering algorithm DADA2 -> ASV's
Read mapping to ASV minimap2, at 0.99 %id cutoff
ASV clustering based on sequence error profiles (-dada2seed 0)
Precluster read merging No
Ref Chimera checking Yes (DB=/usr/local/tools/_conda/envs/__lotus2@2.32/share/lotus2-2.32-0//DB//rdp_gold.fa, -chim_skew 2)
deNovo Chimera check Yes
Tax assignment Lambda (-LCA_frac 0.8, -LCA_cover 0.5, -LCA_idthresh 97,95,88,83,81,78,0)
ReferenceDatabase custom
RefDB location maarjam_database_SSU_TYPE.qiime.fasta
ASV phylogeny Yes (mafft, fasttree2)
Unclassified ASV's Kept in matrix
--------------------------------------------
--------------------------------------------------------------------------------
00:00:03 Reading mapping file
Sequence files are indicated in mapping file.
Found "SequencingRun" column, with 1 categories (a)
--------------------------------------------------------------------------------
--------------------------------------------------------------------------------
00:00:03 Demultiplexing, filtering, dereplicating input files, this
might take some time..
check progress at output/LotuSLogS/LotuS_progout.log
00:02:19 Finished primary read processing with sdm:
Reads processed: 12,127,976
Accepted (High qual): 0 (0 end-trimmed)
Rejected: 12,136,645
Dereplication block 0: 0 unique sequences (avg size -nan; 0 counts)
What this means is that none of your input reads survived the demultiplexing, filtering and dereplication steps, and the most likely reason is that you went with just the default settings for sdm (the tool used internally to carry out that step).
The first lines above tell you that these default settings are the ones from the conda-packaged file /usr/local/tools/_conda/envs/__lotus2@2.32/share/lotus2-2.32-0/configs/sdm_miSeq.txt
, which is for miSeq data and is likely not a good fit for your data.
What you’ll have to do is to provide your own sdm config file like explained in this tutorial. These files may seem complex, but it might be a simple threshold lik e.g. the min read length that’s causing your reads to be filtered out. The meaning of all the settings is explained at LotuS2 — Less OTU scripts for a simpler 16S pipeline.
Hope this helps,
Wolfgang