Skip to content

Multi-sample phasing for ww-glimpse2 and ww-imputation#239

Merged
tefirman merged 5 commits intomainfrom
glimpse2-phase-together
Feb 18, 2026
Merged

Multi-sample phasing for ww-glimpse2 and ww-imputation#239
tefirman merged 5 commits intomainfrom
glimpse2-phase-together

Conversation

@tefirman
Copy link
Member

@tefirman tefirman commented Feb 11, 2026

Type of Change

  • Bug fix
  • Documentation update

Description

Refactors glimpse2_phase_cram and the ww-imputation pipeline to support joint multi-sample phasing, aligning with the Palantir/Broad GLIMPSE2 workflow pattern.

ww-glimpse2 module

  • glimpse2_phase_cram now accepts Array[File] input_bams and Array[File] input_bam_indices instead of a single file. The command block builds a bam_list.txt from the array and passes it to GLIMPSE2_phase via --bam-list, supporting both single-sample and multi-sample use cases.

ww-imputation pipeline

  • Removed the ImputationSample struct and the outer per-sample scatter. All CRAMs are now passed into each glimpse2_phase_cram call for joint phasing across samples.
  • The pipeline now produces a single multi-sample output file instead of one file per sample.
  • Concordance is now a single optional call on the final multi-sample VCF, controlled by workflow-level truth_vcf/truth_vcf_index inputs (instead of per-sample truth VCFs in the struct).
  • Added output_prefix, concat_cpu_cores, and concat_memory_gb parameters.

Related Issue

Testing

How did you test these changes?

What workflow engine did you use?

Did the tests pass?

Documentation

  • I updated the README (if applicable)
  • I added/updated parameter descriptions in the WDL (if applicable)
  • I ran make docs-preview to check documentation rendering (if applicable)

Additional Context

This is a breaking change to the ww-imputation pipeline interface — the ImputationSample struct is removed and replaced with flat Array[File] inputs. The ww-glimpse2 module task glimpse2_phase_cram also has a breaking input signature change (FileArray[File]). Both modules are at prototype status so this is expected.

@tefirman tefirman changed the title Joint multi-sample phasing for ww-glimpse2 and ww-imputation Multi-sample phasing for ww-glimpse2 and ww-imputation Feb 11, 2026
@tefirman tefirman marked this pull request as ready for review February 12, 2026 00:02
@tefirman tefirman requested a review from emjbishop as a code owner February 12, 2026 00:02
@tefirman tefirman merged commit 0fbce71 into main Feb 18, 2026
15 checks passed
@tefirman tefirman deleted the glimpse2-phase-together branch February 18, 2026 05:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Multi-sample phasing in glimpse2_phase_cram

2 participants