NFCORE_RNASEQ:RNASEQ:QUANTIFY_STAR_SALMON:TX2GENE error on iGenomes TAIR10

[holmrenser]

Description of the bug

I tried running nf-core/rnaseq 3.13.2 on the Arabidopsis thaliana TAIR10 genome from iGenomes and ran into an issue with the tx2gene step of processing the annotation gtf. I have used previous pipeline versions on the same genome without this issue.

Command used and terminal output

nextflow run nf-core/rnaseq --input samplesheet_full.csv --outdir mapping_full --genome TAIR10 --max_cpus 8 -profile docker -r 3.13.2

ERROR ~ Error executing process > 'NFCORE_RNASEQ:RNASEQ:QUANTIFY_STAR_SALMON:TX2GENE (genome.filtered.gtf)'

Caused by:
  Process `NFCORE_RNASEQ:RNASEQ:QUANTIFY_STAR_SALMON:TX2GENE (genome.filtered.gtf)` terminated with an error exit status (1)

Command executed:

  tx2gene.py \
      --quant_type salmon \
      --gtf genome.filtered.gtf \
      --quants quants \
      --id gene_id \
      --extra gene_name \
      -o tx2gene.tsv

  cat <<-END_VERSIONS > versions.yml
  "NFCORE_RNASEQ:RNASEQ:QUANTIFY_STAR_SALMON:TX2GENE":
      python: $(python --version | sed 's/Python //g')
  END_VERSIONS

Command exit status:
  1

Command output:
  (empty)

Command error:
  WARNING: Your kernel does not support swap limit capabilities or the cgroup is not mounted. Memory limited without swap.
  Traceback (most recent call last):
    File "/home/<omitted>/.nextflow/assets/nf-core/rnaseq/bin/tx2gene.py", line 162, in <module>
      if not map_transcripts_to_gene(args.quant_type, args.gtf, args.quants, args.id, args.extra, args.output):
    File "/home/<omitted>/.nextflow/assets/nf-core/rnaseq/bin/tx2gene.py", line 122, in map_transcripts_to_gene
      transcript_attribute = discover_transcript_attribute(gtf_file, transcripts)
    File "/home/<omitted>/.nextflow/assets/nf-core/rnaseq/bin/tx2gene.py", line 59, in discover_transcript_attribute
      attributes = dict(item.strip().split(" ", 1) for item in cols[8].split(";") if item.strip())
  ValueError: dictionary update sequence element #4 has length 1; 2 is required

Work dir:
  /lustre/<omitted>/full_experiment/work/57/7fc5bc40a9829787f3723fa27f46a9

Tip: you can replicate the issue by changing to the process work dir and entering the command `bash .command.run`

 -- Check '.nextflow.log' file for details

Relevant files

No response

System information

Nextflow version: 23.10.0.5889
Hardware: 96CPU server, lustre filesystem
Executor: local
Container engine: Docker
OS: Ubuntu 20.04.6
Version of nf-core/rnaseq: 3.13.2

[Guy2Horev]

I got the same error today after the workflow was updated.

[pinin4fjords]

This is genuinely a bad GTF file rather than a pipeline issue: there's a semicolon in one of the gene names, specifically at line 33090

1   ensembl CDS 4810488 4811109 .   +   0   exon_number "1"; gene_biotype "protein_coding"; gene_id "AT1G14040"; gene_name "PHO1;H3"; gene_source "ensembl"; gene_version "1"; p_id "P3587"; protein_id "AT1G14040.1"; protein_version "1"; transcript_biotype "protein_coding"; transcript_id "AT1G14040.1"; transcript_name "PHO1;H3"; transcript_source "ensembl"; transcript_version "1"; tss_id "TSS29975";

"PHO1;H3" is not a good value and it's upsetting the parsing of the semicolon-delimited attributes field.

It wasn't an issue previously because we didn't sample enough lines (which @MatthiasZepper fixed).

I'll try to add something to skip a limited number of bad lines (we don't need them all for this part of the code). In the meantime I recommend you review our guidelines on reference file usage- you really are better off using more recent files from Ensembl (and you can complain to Ensembl about invalid formatting like this).

[pinin4fjords]

Actually, I think I can do better and allow those semicolons- PR incoming. I still maintain they're a silly idea though...

[drpatelh]

Fixed in #1150