Requesting conda package

[sagnikbanerjee15]

Hello,

I was wondering if there is a conda package from where PsiCLASS could be installed.

Thank you.

[mourisl]

Thank you for the suggestion. I just submitted the PsiCLASS recipe to bioconda: https://github.com/bioconda/bioconda-recipes/pull/25986 . I will let you know when they accept the request.

[sagnikbanerjee15]

That's great thanks!

[mourisl]

PsiCLASS is now available at bioconda: https://anaconda.org/bioconda/psiclass !

[palfalvi]

Hi,

Thanks for the package and for the conda package. It would make a lot easier if I could integrate PsiCLASS into automatic pipelines with conda instead of the manual installation of CLASS2 or PsiCLASS. However, at the moment the conda installation does not work on my side. Looks like psiclass can be run:

$ psiclass

Usage: ./psiclass [OPTIONS]
Required:
    -b STRING: paths to the alignment BAM files. Use comma to separate multiple BAM files
        or
    --lb STRING: path to the file listing the alignments BAM files
Optional:
    -s STRING: path to the trusted splice file (default: not used)
    -o STRING: prefix of output files (default: ./psiclass)
    -p INT: number of processes/threads (default: 1)
    -c FLOAT: only use the subexons with classifier score <= than the given number (default: 0.05)
    --sa FLOAT: the minimum average number of supported read for retained introns (default: 0.5)
    --vd FLOAT : the minimum average coverage depth of a transcript to be reported (defaults: 1.0)
    --maxDpConstraintSize: the number of subexons a constraint can cover in DP. (default: 7. -1 for inf)
    --bamGroup STRING: path to the file listing the group id of BAMs in the --lb file (default: not used)
    --primaryParalog: use primary alignment to retain paralog genes (default: use unique alignments)
    --version: print version and exit
    --stage INT:  (default: 0)
        0-start from beginning - building splice sites for each sample
        1-start from building subexon files for each sample
        2-start from combining subexon files across samples
        3-start from assembling the transcripts for each sample
        4-start from voting the consensus transcripts across samples

However when I actually try to run it looks for perl files and the samtools folder, which is not in the conda directory:

$ psiclass -b combined.starAligned.sortedByCoord.out.bam

sh: /.../miniconda3/bin/samtools-0.1.19/samtools: No such file or directory
Found mate read id index suffix(.1 or /1). Calling "--mateIdx 1" option. If this is a false calling, please use "--mateIdx 0".
/.../miniconda3/bin/junc combined.starAligned.sortedByCoord.out.bam -a  --hasMateIdSuffix > ./splice/psiclass_bam_0.raw_splice
/.../miniconda3/bin/trust-splice ./splice/psiclass_splice.list combined.starAligned.sortedByCoord.out.bam  > ./splice/psiclass_bam.trusted_splice
perl /.../miniconda3/bin/FilterSplice.pl ./splice/psiclass_bam_0.raw_splice ./splice/psiclass_bam.trusted_splice > ./splice/psiclass_bam_0.splice
Can't open perl script "/mnt/.../miniconda3/bin/FilterSplice.pl": No such file or directory
Terminated

Do you have any solution for this? Sorry, I am not familiar with conda packages, cannot directly comment on those files.

Thanks.

[mourisl]

I guess conda did not move the non-executable script to the conda bin folder. I'm not familiar with conda either, let me try to repackage psiclass and update the conda receipt. Thank you!