Funcotator outputs fields in wrong order in some cases when writing vcf #6173
Assignees
Comments
lbergelson
changed the title
lbergelson
added a commit
that referenced
this issue
Sep 24, 2019
Fixes #6173 I added a single line test that shows the problem, and regenerated the other test files. I didn't write a specific unit test that proves it's solved. Feel free to do so if you think it's necessary. I validated the output by adding the field name to the output value and checking it by eye against the header lines. This could fairly simply made into a unit test if desired. I'm not sure if there are other large files that need to be regenerated. I had initially said this bug only affects vcf but I think it happens to Maf output as well.
lbergelson
added a commit
that referenced
this issue
Sep 24, 2019
Fixes #6173 I added a single line test that shows the problem, and regenerated the other test files. I didn't write a specific unit test that proves it's solved. Feel free to do so if you think it's necessary. I validated the output by adding the field name to the output value and checking it by eye against the header lines. This could fairly simply made into a unit test if desired. I'm not sure if there are other large files that need to be regenerated. I had initially said this bug only affects vcf but I think it happens to Maf output as well.
lbergelson
added a commit
that referenced
this issue
Oct 3, 2019
Fixes #6173 I added a single line test that shows the problem, and regenerated the other test files. I didn't write a specific unit test that proves it's solved. Feel free to do so if you think it's necessary. I validated the output by adding the field name to the output value and checking it by eye against the header lines. This could fairly simply made into a unit test if desired. I'm not sure if there are other large files that need to be regenerated. I had initially said this bug only affects vcf but I think it happens to Maf output as well.
|
The issue seems to happen when you have multiple records in the datasource which exactly match the position and alleles of a record in the file being annotated. The output fields for that datasource will be scrambled at those sites. This affects both vcf and maf output. |
lbergelson
added a commit
that referenced
this issue
Oct 4, 2019
* Fix VCFFuncotation output order bug
This bug caused funcotation fields produced from a VCFFuncotation source to be scrambled when there were multiple entries with identical positions/alleles in the datasource vcf.
This was caused by accidental use Collectors.toMap creating a HashMap instead of a LinkedHashMap
* Removed a method which allowed passing in a Map instead of LinkedHashMap in order to harden against future bugs of this source
* Fixes #6173
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
There's an issue in
VcfFuncotationFactory.createFuncotationsOnVariantwhere certain cites that need multiple functotations merged produce incorrect output. The cause is an accidental conversion of LinkedHashMap -> HashMap which scrambles the iteration order of the map. The order is used when writing the fields so the names of the fields end up on different values.The text was updated successfully, but these errors were encountered: