Releases: jeremymcrae/clinical-filter
Releases · jeremymcrae/clinical-filter
0.5.0
- restructured into python package
- refactored code
- split out info into separate class
- account for multinucleotide variant consequences
- lowered minor allele frequency threshold
- added imprinting, X-linked over-dominant and mosaic modes of inheritance
- fixed problem loading singletons
- fixed problem with CNVs with missing MADL2R values
0.3.6
- shifted start_lost to loss-of-function consequences
- fixed bug where we missed long CNVs not in known genes
- CNVs now use the per-gene symbols when available
- turned on filtering using ExAC frequencies
- reworked removing variants based on polyphen predictions
- now use per gene polyphen predictions
- all variants with polyphen benign get removed, not just the
ones in trios
- re-include compound hets in proband-only analyses
- fixed bugs that lead to missing DDG2P CNVs under various scenarios
- updated rules for CNVs overlapping syndrome regions
- exclude monoallelic and X-linked dominant variants with high
ExAC allele frequencies - exclude compound hets when proband-only where both hets are
missense - only include compound hets in trios when in trans
- fixed bug when checking if CNVs intersect known gene boundaries
0.3.5
- bug fix for CNV inheritance when analysing proband-only.
- include X-linked over-dominance mode of inheritance.
- fix bug when parsing ped files when only one parent is specified.
- numerous changes to proband-only filtering.
- lower MAF threshold for monoallelic and X-linked dominant variants
when analysing proband-only. - remove heterozygous variants from recessive genes when
analysing proband-only.
- lower MAF threshold for monoallelic and X-linked dominant variants
0.3.4
- minor bug fixes for CNV filtering, and case with multi-allelic sites where the alleles affect different "genes"
- updated paths for new datafreeze in helper scripts
0.3.3
- include CNVs in filter output
- allow for alternate gene symbols to capture additional compound
heterozygotes (useful when not using a list of known genes).
0.3.2
Don't exclude variants that fail VQSLOD threshold, as the variants that pass all other criteria but fail this aspect contain a sufficiently high proportion of reportable variants. This adds around 20% more inherited variants.
0.3.1
minor fix for variants with single consequence, no HGNC, but multiple HGNC_ALL
0.3.0
Allow for per-gene consequences
0.2.0
- exclude inherited male chrX variants with non-zero hemizygous counts in ExAC.
0.1.2
- include additional gene annotations when available
- drop the known genes overlap checking, since we have the additional gene annotations