matsengrp / phip-flow

A Nextflow pipeline to align, merge, and organize large PhIP-Seq datasets
MIT License
10 stars 6 forks source link

Unexpected behavior during nextflow run #66

Closed bkellman closed 1 year ago

bkellman commented 1 year ago

I am running the phip-flow pipeline pancov demo using:

nextflow run matsengrp/phip-flow -r V1.11 -profile docker \
        --sample_table ../code/phip-flow/data/pan-cov-example/sample_table_with_beads_and_lib2.csv \
        --peptide_table ../code/phip-flow/data/pan-cov-example/peptide_table.csv \
        --output_tall_csv true \
        --output_wide_csv true \
        --results "$(date -I)" \
        -resume

While the run terminated with several outputs, there were also multiple errors. nextflow.pancov.test.log.log Is this expected behavior? Is this resulting from an incomplete install?

bkellman commented 1 year ago

I played around a bit overnight and I think I may have resolved by re-installing BEER. Why would that impact alignment completion?

jgallowa07 commented 1 year ago

@bkellman Sorry for the headache here. It's hard to tell what going on with the log file when you're using using inputs which are not provided by the repository (sample_table_with_beads_and_lib2.csv). I'm assuming for now that your version simply updates the paths (fastq_filepath column) in the provided sample table relative to ../code/phip_flow/. In any case, it seems the log file you provided is saying the run completed as expected.

 Workflow completed > WorkflowStats[succeededCount=40; failedCount=0

Can you specify which errors you are referring to? When I replicate this directory structure and run the command you supplied above, I get the following log output:

replicate_run.log

This looks normal/expected to me.

I played around a bit overnight and I think I may have resolved by re-installing BEER. Why would that impact alignment completion?

You are correct, the installation of BEER should not impact the alignment at all. But you're using the --profile docker option, correct? In this case, you should not need to install anything except docker and nexflow. Out publicly hosted containers should handle all dependencies for you.

bkellman commented 1 year ago

@jgallowa07 thanks for the quick response and sorry for the delay. I've been stuck at the bench the last few weeks. Returning to the code, I now realize that not every module is supposed to run. I think that was my mistake.

Thanks for your thorough response and sorry for my mistake.