You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently, we assign Number=. for repeated INFO/FORMAT fields that are not under alternate_bases (i.e. have Number=A). We output . for any field that is null in the BigQuery table, which is fine according to the spec if the Number is also defined as ..
However, if the user provides a --representative_header_file and the value is null, the spec requires the correct number of .s to be present (e.g. if Number=2, then we should use .,.). This gets a bit tricky for Number=G or Number=R as we need to dynamically calculate the correct number depending on the variant.
For fields that are not null, we can use the same approach to validate that correct number of values are set when outputting the VCF lines too.
The text was updated successfully, but these errors were encountered:
Currently, we assign
Number=.
for repeated INFO/FORMAT fields that are not underalternate_bases
(i.e. haveNumber=A
). We output.
for any field that is null in the BigQuery table, which is fine according to the spec if the Number is also defined as.
.However, if the user provides a
--representative_header_file
and the value is null, the spec requires the correct number of.
s to be present (e.g. ifNumber=2
, then we should use.,.
). This gets a bit tricky forNumber=G
orNumber=R
as we need to dynamically calculate the correct number depending on the variant.For fields that are not null, we can use the same approach to validate that correct number of values are set when outputting the VCF lines too.
The text was updated successfully, but these errors were encountered: