SUPPA cluster events error

[paolo-kunderfranco]

Description of the bug

Hi,
I am trying dev version and I encountered the following error:

I have 32 samples total, 4 conditions, 8 samples per condition

Best

Error executing process > 'NFCORE_RNASPLICE:RNASPLICE:SUPPA_SALMON:CLUSTEREVENTS_IOE (CP-Cortex)

Caused by:
  Process `NFCORE_RNASPLICE:RNASPLICE:SUPPA_SALMON:CLUSTEREVENTS_IOE (CP-Cortex)` terminated with an error exit status (1)

Command executed:

  suppa.py \
      clusterEvents \
      --dpsi CP-Cortex_local_diffsplice.dpsi \
      --psivec CP-Cortex_local_diffsplice.psivec \
      --dpsi-threshold 0.05 \
      --eps 0.05 \
      --metric euclidean \
      --min-pts 20 \
      --groups 1-25,6-27,12-30,17-32 \
      --clustering DBSCAN \
        -o CP-Cortex_local_cluster
  
  cat <<-END_VERSIONS > versions.yml
  "NFCORE_RNASPLICE:RNASPLICE:SUPPA_SALMON:CLUSTEREVENTS_IOE":
      suppa: $(python -c "import pkg_resources; print(pkg_resources.get_distribution('suppa').version)")
  END_VERSIONS

Command exit status:
  1

Command output:
  (empty)

Command error:
  ERROR:lib.cluster_tools:Invalid index. Index 32 is larger than the number of columns in the file (16).`

Command used and terminal output

nextflow run nf-core/rnasplice --input $wd/samplesheet.csv \
							   --contrasts $wd/contrast.csv \
							   --outdir splice_attempt1 \
							   --genome GRCm38 \
							   --aligner star_salmon \
							   --min_samps_gene_expr 2 \
							   --min_samps_feature_expr 2 \
							   --min_gene_expr 5 \
							   --min_feature_expr 5 \
							   --min_feature_prop 0.2 \
							   --dexseq_exon true \
							   --edger_exon true\
							   --dexseq_dtu true \
							   --dtu_txi scaledTPM \
							   --rmats true \
							   --rmats_read_len 75 \
							   --suppa true \
							   --suppa_per_isoform true \
							   --save_reference true \
							   --sashimi_plot true \
							   -config config.config \
							   -profile singularity \
							   -r dev

Relevant files

nextflow.log

System information

nextflow/22.10.4
Debian 10
Slurn

[expansion-bioemformatics]

I am having the same issue running via Tower on AWS Batch.

[asmaali98]

Hi @paolo-kunderfranco and @expansion-bioemformatics , thanks for raising the issue. We will work on having this error resolved.

[dkoppstein]

I am also getting a similar issue. I am not sure, but I think the problem occurs because all contrasts are passed to suppa via --groups, whereas the PSI files are somehow already specific for the contrast of interest. Therefore, there are more columns than samples.

[dkoppstein]

To be clear, this happens whenever there are >2 contrasts specified in the samplesheet. I was able to "solve" the problem by running the pipeline multiple times, once for each desired contrast. However, there should be a contrastsheet with multiple contrasts specified in one of the test configs.

[amizeranschi]

Hi @dkoppstein

What do you mean by whenever there are >2 contrasts specified in the samplesheet ?

Did you mean >2 contrasts in the contrast sheet? Or >2 samples in the samplesheet?

I'm also running into this issue, with the following samplesheet and contrast sheet:

sample,condition,strandedness,fastq_1,fastq_2
K562_TARDBP_TRT_REP1,K562_TARDBP_TRT,reverse,fastq-test/SRR3469414-R1.fastq.gz,fastq-test/SRR3469414-R2.fastq.gz
K562_TARDBP_TRT_REP2,K562_TARDBP_TRT,reverse,fastq-test/SRR3469415-R1.fastq.gz,fastq-test/SRR3469415-R2.fastq.gz
K562_TARDBP_CTL_REP1,K562_TARDBP_CTL,reverse,fastq-test/SRR3469418-R1.fastq.gz,fastq-test/SRR3469418-R2.fastq.gz
K562_TARDBP_CTL_REP2,K562_TARDBP_CTL,reverse,fastq-test/SRR3469419-R1.fastq.gz,fastq-test/SRR3469419-R2.fastq.gz
HepG2_TARDBP_TRT_REP1,HepG2_TARDBP_TRT,reverse,fastq-test/SRR4421909-R1.fastq.gz,fastq-test/SRR4421909-R2.fastq.gz
HepG2_TARDBP_TRT_REP2,HepG2_TARDBP_TRT,reverse,fastq-test/SRR4421908-R1.fastq.gz,fastq-test/SRR4421908-R2.fastq.gz
HepG2_TARDBP_CTL_REP1,HepG2_TARDBP_CTL,reverse,fastq-test/SRR4421557-R1.fastq.gz,fastq-test/SRR4421557-R2.fastq.gz
HepG2_TARDBP_CTL_REP2,HepG2_TARDBP_CTL,reverse,fastq-test/SRR4421558-R1.fastq.gz,fastq-test/SRR4421558-R2.fastq.gz

contrast,treatment,control
K562_TARDBP_TRT_CTL,K562_TARDBP_TRT,K562_TARDBP_CTL
HepG2_TARDBP_TRT_CTL,HepG2_TARDBP_TRT,HepG2_TARDBP_CTL

[amizeranschi]

Actually, my error is very similar, but not identical to the one above:

ERROR ~ Error executing process > 'NFCORE_RNASPLICE:RNASPLICE:SUPPA_SALMON:CLUSTEREVENTS_IOI (HepG2_TARDBP_TRT-HepG2_TARDBP_CTL)'

Caused by:
  Process `NFCORE_RNASPLICE:RNASPLICE:SUPPA_SALMON:CLUSTEREVENTS_IOI (HepG2_TARDBP_TRT-HepG2_TARDBP_CTL)` terminated with an error exit status (1)

Command executed:

  suppa.py \
      clusterEvents \
      --dpsi HepG2_TARDBP_TRT-HepG2_TARDBP_CTL_transcript_diffsplice.dpsi \
      --psivec HepG2_TARDBP_TRT-HepG2_TARDBP_CTL_transcript_diffsplice.psivec \
      --dpsi-threshold 0.05 \
      --eps 0.05 \
      --metric euclidean \
      --min-pts 20 \
      --groups 1-2,3-4,5-6,7-8 \
      --clustering DBSCAN \
        -o HepG2_TARDBP_TRT-HepG2_TARDBP_CTL_transcript_cluster
  
  cat <<-END_VERSIONS > versions.yml
  "NFCORE_RNASPLICE:RNASPLICE:SUPPA_SALMON:CLUSTEREVENTS_IOI":
      suppa: $(python -c "import pkg_resources; print(pkg_resources.get_distribution('suppa').version)")
  END_VERSIONS

Command exit status:
  1

Command output:
  (empty)

Command error:
  ERROR:lib.cluster_tools:Invalid index. Index 8 is larger than the number of columns in the file (4).

[dkoppstein]

Sorry -- I did indeed mean >2 contrasts in the contrast sheet. And it is clear that the error is similar, i.e. that all groups are specified on the command line, even though the dpsi and psivec files are already restricted to the contrast specified.

[amizeranschi]

Good catch, I think you are right and the error seems to be caused by how the --groups setting is being specified by the pipeline.

[jma1991]

Thanks for the comments everyone. I'm working on pushing a fix for this today. Hopefully report back with a PR!

[paolo-kunderfranco]

exactly, I did have more contrast specified and the pipelne failed in SUPPA.py

ERROR:lib.cluster_tools:Invalid index. Index 32 is larger than the number of columns in the file (16).

I am trying now to run again with only one contrast specified.

[paolo-kunderfranco]

update, with only one contrast specified, CLUSTEREVENTS_IOE completed succesfully, however I am now facing and error
with NFCORE_RNASPLICE:RNASPLICE:DRIMSEQ_DEXSEQ_DTU_SALMON:STAGER,
I do need to open a new issue?
Best

[tijeco]

I'm not sure if this is helpful, but I've looked into this before and had a similar error as @paolo-kunderfranco ERROR:lib.cluster_tools:Invalid index. Index 32 is larger than the number of columns in the file (16). I noticed this happened if the samplesheet wasn't sorted by condition, because it seems like suppa assumes that all the samples are sorted by condtion.

[jma1991]

Thanks for your patience on this bug. I believe we have fixed the issue with #86. Please can you pull from dev and see if this works.

[jma1991]

Closing because all tests are passing and fix has been merged into dev branch. Please re-open if you still encounter an error.