Working PoC of VCF import API #4498
Conversation
There was a problem hiding this comment.
nice.
I suppose this depends on the data model behind how to handle the VCF contents. So we either should standardize or we should make a map of 'adapters' that can choose how to translate the VCF in the backend (based on the profile that was used to create the schema).
| @@ -0,0 +1,78 @@ | |||
| package org.molgenis.emx2.beaconv2.vcfimport; | |||
|
|
|||
| import htsjdk.variant.variantcontext.VariantContext; | |||
| private Table individuals; | ||
|
|
||
| public VcfImport(File vcfTmpFile, Table genomicVariations, Table individuals) { | ||
| this.vcfTmpFile = vcfTmpFile; |
There was a problem hiding this comment.
would call it 'vcfFile' (might or might not be tempfile, doesn't matter here)
| for (VariantContext variant : vcfFileReader) { | ||
|
|
||
| Row row = new Row(); | ||
| row.set("variantInternalId", UUID.randomUUID()); |
There was a problem hiding this comment.
so this would then need to be an adapter. I.e. using the 'strategy' design pattern where the way on how to handle rows from the VCF file can have different implementations.
There was a problem hiding this comment.
Sounds good, but this needs to be fully transparant to the user so I think we need to make https://github.com/molgenis/GCC/issues/775 into a more general GUI tool
|
|
||
| public static void main(String[] args) { | ||
| String vcfFilePath = | ||
| "/Users/joeri/git/molgenis-emx2/backend/molgenis-emx2-beacon-v2/src/test/resources/sample.vcf"; |
There was a problem hiding this comment.
Ofcourse, this was just for playing around with HTSJDK
|
Started writing VCF import API, got very minimal PoC working using
But many things still to do