Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Anita Schurch
    @aschuerch_gitlab
    Is it okay if I add more people?
    Anita Schurch
    @aschuerch_gitlab
    @SilasK
    Silas Kieser
    @SilasK
    Yes, of course.
    MalbertR
    @MalbertR
    Hi Silas, first, thanks for setting this up!
    I have a question regarding EggNogmapper...
    It seems to be the limiting tool in the pipeline for me at the moment.
    It apparently needs a huge amount of memory. Have you also experienced this before?
    FYI, my 5 subsets.faa contain about 480k genes each
    MalbertR
    @MalbertR
    I went and had a better look at what was going on during the running of EggNogmapper and it seems like the job is not being submitted to compute nodes (like is the case when running the assembly for example), but is ran "locally". Is there a specific reason for this? Could this be the problem?
    Silas Kieser
    @SilasK
    I know eggnog uses a lot of resources. I can not make the tool better than it is, even there might soon be a new release of eggNOG mapper.
    No eggNOG mapper shouldn't run locally. there are some rules before that run locally. But only for filter_genes I saw problems when running millions of genes.
    There are many optimizations for eggnong mapper:
    1. split the genes to annotate into smaller subsets
    1. use virtual memory for annotation
    MalbertR
    @MalbertR
    I'm asking about the running locally since I do not see the submit command like I see it for the other tools
    For example:

    /hpc/dla_mm/mrogers/metagenomics/atlas_db/conda_envs/fecb35ac/bin/diamond /hpc/dla_mm/mrogers/metagenomics/atlas_db/conda_envs/fecb35ac/lib/python2.7/site-packages

    emapper-2.0.1

    ./emapper.py -m diamond --no_annot --no_file_comments --data_dir /hpc/dla_mm/mrogers/metagenomics/atlas_db/EggNOGV2 --cpu 1 -i Genecatalog/subsets/genes/subset2.faa -o Genecatalog/subsets/genes/subset2 --override

    /hpc/dla_mm/mrogers/metagenomics/atlas_db/conda_envs/fecb35ac/bin/diamond blastp -d /hpc/dla_mm/mrogers/metagenomics/atlas_db/EggNOGV2/eggnog_proteins.dmnd -q /hpc/dla_mm/mrogers/metagenomics/Roos/fecal/metagenome-atlas/Genecatalog/subsets/genes/subset2.faa --more-sensitive --threads 1 -e 0.001000 -o /hpc/dla_mm/mrogers/metagenomics/Roos/fecal/metagenome-atlas/.snakemake/shadow/tmpoqbrrhm9/emappertmp_dmdn_ZyRLbj/655d9b5045a74232b30e29af227ec7e9 --top 3 --query-cover 0 --subject-cover 0

    Doesn't look to good in gitter, but that's what I see when running Eggnog
    Silas Kieser
    @SilasK
    Let's continue discussing here: metagenome-atlas/atlas#351
    MalbertR
    @MalbertR
    Yes, let's do that. Thanks.
    Silas Kieser
    @SilasK
    You can also use code
    MalbertR
    @MalbertR
    Ah, didn't know code also worked here...