Split_Multi vs Split_Multi_HTS

I figured out that the example code provided by the platform I am using used split_multi, instead of split_multi_hts, which caused problems with exporting to bgen, giving the following error:

Error summary: HailException: found allele outside of expected range [0, 2]: 3

I already processed two data structures using split_multi, producing bgen files, that did not give me this error. Do I need to reproduce those data structures with split_multi_hts, as split_multi perhaps does not update relevant data fields?

Hi @jwillett,
Yes, the data you’ve already processed with split_multi will have some fields which need to be updated. But that doesn’t require running split_multi_hts, which would redo all the work already done by split_multi. If you look at the example in the documentation for split_multi, you can see the implementation of split_multi_hts, which just does a simple step after split_multi. You should only need to run that second step yourself.

Sounds great, thanks! I saw that code. Glad I do not have to redo much!