build error due to empty fastq files

[michoug]

Hi
When trying to run the command

binchicken build --conda-prefix /work/river/Software/miniconda3/envs

I got this error (the full log is attached binchicken.log)

Using shell: /usr/bin/bash
Provided cores: 1 (use --cores to define parallelism)
Rules claiming more threads will be scaled down.
Job stats:
job               count
--------------  -------
all                   1
aviary_combine        1
cluster_graph         1
count_bp_reads        1
summary               1
total                 5

Select jobs to execute...

[Thu Aug 22 13:56:44 2024]
rule count_bp_reads:
    input: /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_1_unmapped.1.fq.gz, /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_2_unmapped.1.fq.gz, /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_3_unmapped.1.fq.gz, /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_1_unmapped.2.fq.gz, /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_2_unmapped.2.fq.gz, /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_3_unmapped.2.fq.gz
    output: /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/unmapped_read_size.csv
    jobid: 30
    reason: Missing output files: /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/unmapped_read_size.csv
    wildcards: version=unmapped_
    resources: tmpdir=/tmp, mem_mb=8000, mem_mib=7630, runtime=1440


gzip: /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_1_unmapped.1.fq.gz: unexpected end of file

gzip: /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_2_unmapped.1.fq.gz: unexpected end of file

gzip: /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_3_unmapped.1.fq.gz: unexpected end of file
[Thu Aug 22 13:56:44 2024]
Finished job 30.
1 of 5 steps (20%) done
Select jobs to execute...

The files /scratch/gmichoud/BinChicken/build/build_aviary/coassemble/mapping/sample_1_unmapped.1.fq.gz and other are empty
In the directory where I run this command are present those files all empty

-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 genome_1.fna
-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 sample_1.1.fq
-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 sample_1.2.fq
-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 sample_2.1.fq
-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 sample_2.2.fq
-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 sample_3.1.fq
-rw-r--r-- 1 gmichoud river-unit    0 Aug 22 13:56 sample_3.2.fq

Any ideas
Best
Greg

[AroneyS]

This is a bug. Adding a fix now, but note that binchicken is installed fine. This will only delay the running of assembly/binning through Aviary with the --run-aviary flag.

[AroneyS]

Actually I can't replicate this. Is there a build/build_aviary/coassemble/summary.tsv file? What version of binchicken are you using?

[michoug]

Hi
It's version 0.12.1 and there is a build/build_aviary/coassemble/summary.tsv but it's empty
Greg

[michoug]

Also @AroneyS, for some reasons, it works on an Ubuntu 20.04.6 LTS computer but not on a Red Hat Enterprise Linux 9.0 HPC cluster

[AroneyS]

The files are all supposed to be empty, they are just placeholders so that snakemake can recognise the conda env that need to be built. Snakemake uses mtime to decide if jobs need to be rerun (e.g. if an input to a rule has a newer file than the output). Maybe there is an issue with file creation being delayed?