Vembrane

Changes

- Format containment fix tedil (90)

Changes

- Add --version option tedil (88)

Features

- Add `INDEX` to available fields for filter/table christopher-schroeder (89)

Changes

- More custom VEP annotation types tedil (84)
- Small performance improvement for `keep_unmatched` tedil (83)
- make FILTER a list tedil christopher-schroeder (82)


We also added function families for working with genotypes (`FORMAT["GT"]`): `count_hom`, `count_het`, `count_any_ref`, `count_any_var`, `count_hom_ref`, `count_hom_var` and `is_hom`, `is_het`, `is_hom_ref`, `is_hom_var`, `has_ref`, `has_var`. These are similar in semantics to the `countHom` etc functions found in SnpSift with the exception that unknown genotype information (i.e. `.` in VCF notation) is **not** considered to be `REF` (i.e. `0` in VCF notation) but... unknown. Consider the following VCF snippet with 27 triploid samples in all possible combinations of `0` (ref) allele, `1` (first alternative) allele and `.` (missing information):


CHROM POS ID REF ALT QUAL FILTER INFO FORMAT S1 S2 S3 S4 S5 S6 S7 S8 S9 T1 T2 T3 T4 T5 T6 T7 T8 T9 U1 U2 U3 U4 U5 U6 U7 U8 U9
chr18 27963423 . G A 276 PASS GT ././. 1/1/. 0/0/. 0/1/. 1/0/. ./0/. ./1/. 1/./. 0/./. ././0 1/1/0 0/0/0 0/1/0 1/0/0 ./0/0 ./1/0 1/./0 0/./0 ././1 1/1/1 0/0/1 0/1/1 1/0/1 ./0/1 ./1/1 1/./1 0/./1

vembrane reports the following counts:
- `count_hom() == 2` (only `0/0/0` and `1/1/1` are considered homozygous)
- `count_het() == 12`
- `count_any_ref() == 19`
- `count_any_var() == 19`
- `count_hom_ref() == 1`
- `count_hom_var() == 1`

SnpSift reports the following counts:
- `countHom() == 3` (`././.` is considered homozygous in addition to `0/0/0` and `1/1/1`)
- `countHet() == 24`
- `countRef() == 8` (since `.` is considered `0`)
- `countVariant() == 19`

Changes
- `vembrane "expression" input` is now `vembrane filter "expression" input`

Features

- vembrane now supports filtering VCF files with breakends (BNDs). This requires two passes if you wish to keep the order of the input, which can be enabled with the `--preserve-order` option. christopher-schroeder (80)
- since we already evaluate expressions, we might as well use this to write tabular output: `vembrane table "expression" input` evaluates the given expression and writes tabular data as tsv, check `vembrane table --help` for more information christopher-schroeder (79)

Changes

- Fix error if there are no INFO/FORMAT entries; add test christopher-schroeder tedil (77)

Changes

- Only allow one ALT allele per record, error if more than one is encountered tedil (73)
*This change affects expression syntax and VCF input requirements*:
- VCF files need to be split such that each record has only one ALT allele, this can be achieved with e.g. the following tools:
- [`bcftools norm -m-any […]`](http://samtools.github.io/bcftools/bcftools.html#norm)
- [`gatk LeftAlignAndTrimVariants […] --split-multi-allelics`](https://gatk.broadinstitute.org/hc/en-us/articles/360037225872-LeftAlignAndTrimVariants)
- [`vcfmulti2oneallele […]`](http://lindenb.github.io/jvarkit/VcfMultiToOneAllele.html)
- This makes filter expressions way less verbose and easier to use; instead of `ALT[0] == "C" and INFO["gnomad_AF"][0] > 0.01` (where `0` is the index of the first ALT allele), use `ALT == "C" and INFO["gnomad_AF"] > 0.01`. (This does reduce expressiveness a bit since you cannot reason about, say, the sum of alternate allele frequencies anymore)
- (Also consider splitting your VCF files *before* annotating with snpeff/vep/etc.)
- Export NA to enable checks such as `INFO["DP"] is NA` tedil (74)
- bool(NA) = False christopher-schroeder (72)