search

databio / peppro

A modular, containerized pipeline for PRO-seq data processing
http://peppro.databio.org/
BSD 2-Clause "Simplified" License
10 stars 2 forks source link

Error: Unable to access jarfile picard #10

Closed nsheff closed 5 years ago

nsheff commented 5 years ago 

### Merge/link and fastq conversion:  (07-19 20:16:47) elapsed: 0.0 _TIME_

Number of input file sets: 1
Found lock file: /project/shefflab/processed/ppqc/results_pipeline/GSM1558746/lock.raw__GSM1558746.merged.bam
Overwriting target. . .
Target to produce: `/project/shefflab/processed/ppqc/results_pipeline/GSM1558746/raw/GSM1558746.merged.bam`

> `java -Xmx15200M -jar picard MergeSamFiles INPUT=/project/shefflab/data/sra_bam/SRR1693611.bam INPUT=/project/shefflab/data/sra_b
am/SRR1693612.bam OUTPUT=/project/shefflab/processed/ppqc/results_pipeline/GSM1558746/raw/GSM1558746.merged.bam ASSUME_SORTED=TRUE 
CREATE_INDEX=TRUE VALIDATION_STRINGENCY=SILENT` (257)

> `java -Xmx15200M -jar picard MergeSamFiles INPUT=/project/shefflab/data/sra_bam/SRR1693611.bam INPUT=/project/shefflab/data/sra_bam/SRR1693612.bam OUTPUT=/project/shefflab/processed/ppqc/results_pipeline/GSM1558746/raw/GSM1558746.merged.bam ASSUME_SORTED=TRUE CREATE_INDEX=TRUE VALIDATION_STRINGENCY=SILENT` (199)

<pre>
Error: Unable to access jarfile picard
nsheff commented 5 years ago

I tried adding 

tools:  # absolute paths to required tools
  picard: ${PICARD}

and it says:

Changed status from initializing to running.

Loading config file: /home/ns5bc/code/peppro/pipelines/peppro.yaml

The following required tool(s) are not callable: ${PICARD} ${PICARD}
Starting cleanup: 0 files; 0 conditional files for cleanup
nsheff commented 5 years ago

this is using the singularity container. I'm not sure why it wouldn't work for me.

nsheff commented 5 years ago

picard is not actually even listed as a required tool. I think this is coming from ngstk?

nsheff commented 5 years ago

if you're assuming that everything under tools is required, this is evidence that this is a bad idea

jpsmith5 commented 5 years ago

Okay, yes, this is because starting from a BAM file requires picard. I pretty much exclusively have been testing with fastq files as starting point, so I didn't remember that picard is a required tool for ngstk/pypiper then. It's not very clearly indicated anywhere that the use of a subset of ngstk tools does require picard. Not sure just yet whether simply adding the additional required tool is the way to go or if modifying the merge_or_link usage in the pipeline is the better route.

jpsmith5 commented 5 years ago

Added picard as required tool for now.