We've marked this release as the first full release of v2, linked with our recent publication describing the calculator in full (Lambert, Wingfield, et al. Nature Genetics. 2024.).
Changelog
Improvements
- Make report shareable by default
- Remove individual level data
- Don't show density plots with small sample sizes
- Add warnings about complex alleles (e.g. HLA/APOE) and dosage specific effect weights to the report
- The variant verification step added in
2.0.0-beta.3has been integrated intopgscatalog-aggregate- The symmetric difference of scoring file variant IDs (
.scorefile.gz) and variants that contributed to the final calculated score (.varsplink file) must be an empty set
- The symmetric difference of scoring file variant IDs (
Bug fixes
- Stop crashing when encountering a scoring file with dosage specific effect weights (skip instead)
- Fix report logo
- Fix VCF input with JSON samplesheets
- Add tar to zstd conda environment to prevent very old tar installs failing to extract the database
- Reduce download max thread workers to prevent throttling by the EBI
- Fix variant verification step failing in some conda deployments
Full Changelog: v2.0.0-beta.3...v2.0.0
Note
- The
COMBINE_SCOREFILESprocess may take longer to finish and use more memory than in previous versions - New internal variant data models were added in this release to improve handling complex alleles and dosage specific effect weights
- Also, every variant (scoring file row) has many more validation steps now to ensure data quality and consistency
- Speed and memory usage will be improved in the next release