Module installations

[mortonjt]

• [ ] bin
• [x] assemble
• [ ] function
• [ ] map
• [x] raw
• [ ] mash
• [ ] anvio

[sjanssen2]

I am listing external programs I found in the different modules here: I figure native unix tools like "scp", "touch", ... will be there, but we should maybe test their presence, therefore I listed them, too. @tanaes could you double check if these lists are complete / correct?

• anvio
  • [ ] anvi-gen-contigs-database
  • [ ] scp
  • [ ] anvi-run-hmms
  • [ ] anvi-get-dna-sequences-for-gene-calls
  • [ ] centrifuge
  • [ ] anvi-import-taxonomy
  • [ ] anvi-profile
  • [ ] anvi-merge
  • [ ] anvi-import-collection
  • [ ] anvi-summarize
  • [ ] tar
• assemble
  • [ ] megahit
  • [ ] scp
  • [ ] spades.py
  • [ ] ln
  • [ ] metaquast.py
  • [ ] touch
  • [ ] quast.py
  • [ ] multiqc
• bin
  • [ ] samtools
  • [ ] pileup.sh
  • [ ] awk
  • [ ] grep
  • [ ] scp
  • [ ] run_MaxBin.pl
  • [ ] tar
  • [ ] rm
• map
  • [ ] bowtie2-build
  • [ ] bowtie2
  • [ ] samtools
  • [ ] scp
• mash
  • [ ] cat
  • [ ] seqtk
  • [ ] mv
  • [ ] sketch
  • [ ] mash
  • [ ] sort
• raw
  • [ ] fastqc
  • [ ] touch
  • [ ] multiqc
  • [ ] cat
• report
  • [ ] seaborn
  • [ ] pandas
  • [ ] json
  • [ ] yaml
  • [ ] matplotlib
  • [ ] numpy

[tanaes]

Anvio: @mortonjt has been working on a unified conda recipe for this tool, so let's defer on that for now.

I don't think it's that important to test for basic *nix tools, is it?

Assemble: quast.py and metaquast.py are in same package

Bin: pileup.sh is in the BBmap recipe maxbin2 is the other one we need. I have had trouble with this one.

Map: bowtie2-build and bowtie2 are both in bowtie2

Mash: waiting for @RNAer to update with SourMash

Report: I think these actually need to be added to the base oecophylla environment, because these are executed in that context?

[sjanssen2]

I feel that there is a lot of code duplication for all those more or less small .yaml and .sh files for the conda installation of the modules plus a lot of places where they can get out of sync:

• conda env prefix: "oecophilla" vs "shotgu"
• different conda package versions: "hmmer" vs. "hmmer ==3.1."

Could we somehow centralize the knowledge about the modules and generate the necessary .yaml and .sh files?

[RNAer]

the versions can be different in different conda env because one pkg requires v2 and another v3?

On Sep 14, 2017, at 9:37 AM, Stefan Janssen notifications@github.com wrote:

I feel that there is a lot of code duplication for all those more or less small .yaml and .sh files for the conda installation of the modules plus a lot of places where they can get out of sync:

conda env prefix: "oecophilla" vs "shotgu" different conda package versions: "hmmer" vs. "hmmer ==3.1." Could we somehow centralize the knowledge about the modules and generate the necessary .yaml and .sh files?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/biocore/oecophylla/issues/14#issuecomment-329539320, or mute the thread https://github.com/notifications/unsubscribe-auth/AA9CeC8Wp6ES7W4UuhbCIu9Vnujnf7Wpks5siVZWgaJpZM4PVZ3k.

[adswafford]

seqtk does not appear to be installing correctly. If you install following the readthedocs tutorial and then run the test you get:

"/bin/bash: line 8: seqtk: command not found /bin/bash: line 8: seqtk: command not found Error in rule taxonomy_shogun: jobid: 0 output: results/taxonomy/sample_S22282/shogun/sample_S22282.profile.txt log: results/taxonomy/logs/taxonomy_shogun.sample=[sample_S22282].log

RuleException: ..."

Manually installing seqtk and adding it to PATH resolved this issue.