Skip to content

investigate: confirm correct GTF input for ribotish quality in hybrid annotation mode #162

Description

@pinin4fjords

Summary

Unlike ribotish predict, the ribotish quality subcommand does not have a -a secondary annotation flag (confirmed by source code inspection of src/run/quality.py). It accepts only a single -g gene annotation file. This issue determines the correct behaviour when --extended_orf_analysis is active and a hybrid GTF (canonical + novel StringTie transcripts) is available.

Background

ribotish quality is a QC/calibration step: it assesses P-site signal quality by measuring reading-frame periodicity against known coding sequences. Its output drives sample-quality decisions, not ORF discovery.

Option A (likely correct) — always pass canonical annotation only
Novel intergenic transcripts (no CDS) contribute zero signal to frame-periodicity metrics and may dilute QC plot readability. This is consistent with the decision made for riboWaltz (canonical-only permanently).

Option B — pass hybrid GTF
Maintains a single GTF path but provides no benefit: CDS-absent transcripts are invisible to frame-periodicity metrics.

Tasks

  1. Run ribotish quality with a mixed GTF (CDS-bearing + CDS-absent novel transcripts) and check whether QC metrics change relative to canonical-only input.
  2. If Option A is confirmed: document that ribotish quality always uses canonical annotation regardless of --extended_orf_analysis; add an explicit override in conf/modules.config if the pipeline currently passes the wrong GTF.
  3. Close with findings — likely no module change needed, just pipeline config.

References

Metadata

Metadata

Assignees

Labels

questionFurther information is requested

Type

No type

Fields

No fields configured for issues without a type.

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions