v1.8

Added

• Added --flow rapid option for using https://github.com/ZhiGroup/RaPID as a main IBD inference tool for phased datasets
• Custom preprocessing for --flow rapid and --flow germline-king which does not lose any phase information
• simbig workflow for generation of simulated datasets up to 500K samples
• Parallel ersa relationship inference

Changed

• Updated snakemake to the 7.30.2
• Updated other packages
• Reduced docker image size from 7.5GB to 4.5GB by merging conda environments into one
• Fast polars-based IBD segments and relationship postprocessing in postprocess_ersa.py

Fixed

• Added leftover memory flag so that it did not crash with MemoryError
• Fixed error in outlier detection

v1.7

Added

• All conda environments are now built in Dockerfile and Snakemake doesn't need to create them for every workflow run from .yaml files.
• Multiple tests were added, covering all GRAPE flows. Test cases are stored at grape/test-cases.

Changed

• Phased affymetrix chip is now stored within the bundle to speed up the simulation flow, because of this intersect rule in pedsim simulation workflow was moved to reference downloading workflow.

Fixed

• Fixed ibis detecting empty IBD segments causing pipeline teardown.

v1.6.1

Fixed

• Fixed --use-bundle flag usage.
• Updated token to download reference bundles and build dependencies.
• Updated snakemake version to fix the bug: snakemake/snakemake#1899.

v1.6

Added

• New workflow for simulation of a big relatives dataset (~500k samples) was added. It's available via simbig command of the pipeline launcher.
• Support multiple cores for the preprocessing (preprocess) workflow.
• IBD segments weighting feature was added, see compute-weight-mask workflow and --weight-mask parameter of the pipeline launcher.
• Several options for better control of the samples filtering were added: --missing-samples, --alt-hom-samples, --het-samples.
• Random seed parameter was added for the Ped-sim simulation.

Changed

• GRAPE flows were renamed in the pipeline launcnher: ibis_king -> ibis-king, germline -> germline-king.
• readme.md and the GRAPE scheme were updated and actualized.
• Singularity support was removed in favour of conda environments.
• Code refactoring and clean up.

Fixed

• Fixed germline-king simulation flow.
• Fixed java command not found during the reference workflow evaluation.

v1.5.2

1. With --flow ibis_king grape now calculates IBD1 and IBD2 shares from KING data for the 0-3 degrees.
2. Fixed a bug with parsing ersa output for large datasets.
3. Fixed a bug with setting every values for some rows in relatives.tsv to 2.
4. Fixed total_seg_len and total_seg_len_ibd2 calculation. Now total_seg_len corresponds to only ibd1 segments.

v1.5.1

1. Bundle downloading hotfix.
2. File verification hotfix for reference workflow.

v1.5

1. Removed singularity from all workflows.
2. Many intermediate files are now temporary. This significantly reduces working folder size.
3. Fixed removal of duplicate SNPs.
4. ERSA-only workflow now correctly detects duplicates or monozygotic twins.

Dockstore Release

1. Dockstore support
2. Small and full bundle reference downloading from azure
3. ersa can handle 100k samples
4. Preprocessing saves phasing information in vcf input
5. MAF filter is now consistent across different inputs