#17 ✓invalid
john.mccallum (at plantandfood)

BLAST template confusion

Reported by john.mccallum (at plantandfood) | April 4th, 2011 @ 01:59 AM | in v1.0

Creating a new blast job inside a directory calling data from NCBI leads to apparent inheritance of variables from higher level directory...
ie (blast) db is set to NCBI db name...

[Cfljam@pgenome 20.NCBI_GSS]$ pwd /data/moa/db/onion/20.Genbank_onion/20.NCBI_GSS [Cfljam@pgenome 20.NCBI_GSS]$ moa show db nucgss
postcommand
precommand
project
query allium cepa[organism]
sequence_name gss_out
title retrieve onion GSS sequences
[Cfljam@pgenome 20.NCBI_GSS]$ mkdir 10.BLASTvsOnionnt [Cfljam@pgenome 20.NCBI_GSS]$ cd 10.BLASTvsOnionnt/ [Cfljam@pgenome 10.BLASTvsOnionnt]$ moa show Error: This command must be executed in a Moa job directory
[Cfljam@pgenome 10.BLASTvsOnionnt]$ moa new blast -t 'Blast onion GSS vs Onion nt' Created a Moa blast job
[Cfljam@pgenome 10.BLASTvsOnionnt]$ moa show db nucgss
eval 1e-10
gff_blasthit F
gff_source BLAST
input /.fasta
nohits 50
nothreads 2
outgff gff/.gff
output out/
.out
postcommand
precommand
program blastn
project
query allium cepa[organism]
sequence_name gss_out
title Blast onion GSS vs Onion nt

Comments and changes to this ticket

  • MF

    MF April 4th, 2011 @ 03:40 AM

    I agree - that is somewhat confusing... need to think about that one..

    I would typically organize such jobs in leaf nodes:

    ./10.download.from.ncbi ./20.create.blastdb ./30.blast

    but your approach is no less valid ...

  • MF

    MF April 6th, 2011 @ 02:48 AM

    Think I have a solution - make only those vars inheritable that are either defined as being inheritable or NOT in the template definition.

    Thus: 'db' would be in the template definition of blastdb and is not inherited by jobs in subdirs. Any dummy var would be inherited.

  • MF

    MF April 28th, 2011 @ 12:39 AM

    • State changed from “new” to “invalid”

    fixing this is weird - it makes thinks more complicated than need be -

    I suggest that in templates we'll try to use better names for parameters -

    and, possibly, that one should not organise a pipeline in such a way that there are many jobs in leaf directories of other jobs. ALthough this is a matter of taste.

    It will be improved partly by solving ticket [19] - as that will make sure that it is clear if a parameter is inherited.

Please Sign in or create a free account to add a new ticket.

With your very own profile, you can contribute to projects, track your activity, watch tickets, receive and update tickets through your email and much more.

New-ticket Create new ticket

Create your profile

Help contribute to this project by taking a few moments to create your personal profile. Create your profile »

lightweight workflows for bioinformatics

Shared Ticket Bins

People watching this ticket

Pages