Skip to content

vgci tests use randomized simulated reads, but this can make them fail spuriously #4963

Description

@adamnovak

As noted in #4861 (comment), it's possible for the CI tests to simulate reads that BWA-MEM has an unusually difficult time mapping (including reads that end up mismapped at MAPQ 60).

This can make the CI tests fail for reasons unrelated to the branch under test.

We should figure out how to pass a seed through toil-vg for read simulation, and do that, so the tests are deterministic. I don't think, in the last few years, we've ever had those randomized tests fail in a way that let us actually find something to fix.

If it's already seeded to be deterministic, we should figure out why it isn't actually deterministic and fix it.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions