Missing colour and hover labels in interactive report volcano plot

[ajeffs]

Description of the bug

With test data, volcano plots in interactive reports show differential genes in orange. With my own data, the workflow completes without error, but, as shown in the following plot:

1. All data points in the volcano plot are grey in the html and shinyngs reports.
2. The "Differential Status" legend lists "unselected rows".
3. There are no labels on hover.

In contrast, the png in the plots folder does show differentially expressed genes as orange with appropriate legend:

Perhaps this has something to do with the GTF provided for annotation? I note my GTF uses "gene" not "gene_name" column but specifying "gene" with --features_name_col and --differential_feature_name_column makes no difference.

Any advice very much appreciated! Thanks for providing such a great resource!
Regards,
Aaron.

Command used and terminal output

nextflow run nf-core/differentialabundance -r 1.5.0 -profile rnaseq,singularity --input c2bbe1-samplesheet-multiple-concs.csv --contrasts c2bbe1-contrasts-multiple-concs.csv --matrix c2bbe1cf.salmon.merged.gene_counts.tsv --transcript_length_matrix c2bbe1cf.salmon.merged.gene_lengths.tsv --gtf GCF_000002285.5_Dog10K_Boxer_Tasha_genomic.gtf --outdir nf-core_diffabun_v1.5.0_multi-contrast

Output is in the attached nextflow.log

Relevant files

nextflow.nfcda.v1.50.log

System information

• Nextflow version: 23.10.1
• Hardware: HPC
• Executor: slurm
• Container engine: Singularity
• OS: Rocky Linux 9.2 (Blue Onyx) Linux 5.14.0-284.18.1.el9_2.x86_64
• Version of nf-core/differentialabundance: 1.40 and 1.50

[ajeffs]

Follow-up observations in the shiny app:

1. "Assay data" > "Heatmaps": lack of annotation, numbers are shown not gene names/ID.
2. "Gene Info" Gene expression bar plot and contrast table do not load: "waiting for ID list" is displayed in the bar plot and contrast table area, and the gene name doesn't populate in the "Gene" panel.
3. "Differential" > "Differential set intersection": the plots show, but the table doesn't populate "Showing 0 to 0 entries (filtered from NaN total entries".

I'm guessing this and the volcano plot issue are probably all related to the same underlying glitch?

Aaron.

[WackerO]

Hello @ajeffs, I stumbled over the same problem some days ago. However, for me, it was quite easily solvable by setting --differential_feature_name_column. Is it possible you did not set the param correctly? Alternatively, could you try --differential_feature_id_column?

[ajeffs]

Hi @WackerO thanks for the suggestion! I had tried both options before posting here, and neither solved the issue for me.

[WackerO]

Hmm, that's unfortunate :'D @ajeffs Do you by any chance have a small dataset with which I could use to reproduce the error?