Accept enriched genotypes at GENOTYPE filter.

[j-coll]

Genotype filter accepts a list of genotypes per sample. Querying multi-allelic genotypes is not always as easy, because there are multiple combinations that depends on the number of secondary alternates (0/1, 0/2, 1/2, 3/5, ...)

This operation can be pretty easy. Since #192, a list with all the loaded genotypes is stored and updated in the StudyConfiguration.

Option A

We can allow wildcards (*) as "any allele different than 0 or missing". Example:

• 0/* -> 0/1 , 0/2 , 0/3 ... 0/n
• 1/* -> 1/1 , 1/2 , 1/3 ... 1/n
• */* -> 1/1 , 1/2 , 2/2 ...

Option B

An alternative can be accept an n instead instead of a wildcard * but with the same meaning:

• 0/n
• n/n

Option C

There is also other alternate to accept genotype aliases, which will help with polyploidy and phased genotypes:

• HOM_REF -> 0, 0/0, 0/0/0, ...
• HOM_ALT -> 1, 1/1, 1|1, 2/2, 3/3, 1/1/1, ...
• HET -> 0/1, 1/2, 0/2, 2/4, 0|1, 1|0, 0/0/1, ...
• HET_REF -> 0/1, 0/2, 0/3, 0|1, ...
• HET_ALT -> 1/2, 1/3, 2/4, 2|1, ...
• MIS -> ., ./., ././., ...

Conclusion

We will go for Option C, as is the most flexible solution

Tasks

• [x] Define genotype classes
• [x] Allow new genotypes in storage-mongodb
• [x] Allow new genotypes in storage-hadoop
• [x] Add tests
• [x] Improve documentation to add new accepted values

[imedina]

Nice, let's implement the * for now.

[jpflorido]

Although it shouldn't be a problem, be aware that GATK represents spanning deletions as * https://gatkforums.broadinstitute.org/gatk/discussion/6926/spanning-or-overlapping-deletions-allele

[imedina]

That should be a problem (in theory), @j-coll actually the two options to implement are:

• 0/* -> 0/1 , 0/2 , 0/3 ... 0/n
• */* -> 1/1 , 1/2 , 2/2 ...

1/* would be part of */*

[j-coll]

Implementing Option C with new genotypes HOM_REF, HOM_ALT, HET, HET_REF, HET_ALT and MISS