pgsc_calc v2.0.0-alpha.6

Changelog

2024-05-28 update: We're investigating unexpected pgscatalog.core.lib.pgsexceptions.MatchRateError in some environments (e.g. UK Biobank on a HPC). This release has been downgraded to a pre-release

Please note the minimum required nextflow version has been updated to v23.10.0, released in October 2023. Run nextflow self-update to upgrade your nextflow version.

Improvements

• Migrate our custom python tools to new pygscatalog packages
  • Reference / target intersection now considers allelic frequency and variant missingness to determine PCA eligibility
  • Downloads from PGS Catalog should be faster (async)
  • Packages are now documented
• Update plink version to alpha 5.10 final #179
• Add docs describing cloud execution
• Add correlation test comparing calculated scores against known good scores
• When matching variants, matching logs are now written before scorefiles to improve debugging UX
• Improvements to PCA quality (ensuring low missingness and suitable MAF for PCA-eligble variants in target samples).
  • This could allow us to implement MAF/missingness filters for scoring file variants in the future.

Bug fixes

• Fix ancestry adjustment with VCFs #252
• Fix support for scoring files that only have one effect type column #280
• Fix adjusting PGS with zero variance (skip them) #283
• Check for reserved characters in sampleset names

Known bug

• Incorrectly adjusting the AVG in --run_ancestry mode #301
• unexpected pgscatalog.core.lib.pgsexceptions.MatchRateError in some environments (e.g. UK Biobank on a HPC)