v0.9.21-12-g92eb53a failing with -t option

[jpuritz]

Hey Erik,

All of my FreeBayes runs are now failing after a few variants when using the -t option. Something changed between v0.9.21-12-g92eb53a and v0.9.20-4-g42d6a54

[jpuritz]

Any version of 0.9.21 fails.

[ekg]

Is there a quick way I could reproduce this? I'm trying to add test cases,
but I already am testing --targets so I'll need another example to add.

On Tue, Apr 14, 2015 at 5:37 AM, jpuritz notifications@github.com wrote:

Any version of 0.9.21 fails.

—
Reply to this email directly or view it on GitHub
#163 (comment).

[jpuritz]

Erik,

Thanks for the quick reply. I've linked a test case with output and log files from runs with and without targets using the latest version and version 0.9.20.

Thanks!

https://www.dropbox.com/s/r4znronfga5zujr/testcase.tar.gz?dl=0

[jpuritz]

Hey Erik,

I took a quick look at the fix and it seems to be working correctly. I'm getting slightly different results from the two different versions ( on a small test data set) in terms of genotypes and variant calls, but it is somewhere over 99% concordance. The discordance is highest, it seems, at sites that are INDEL or complex at the end of a reference contig. Not too surprising. I'll keep looking into this next week and will keep you informed.

Thanks,

Jon

[ekg]

Hey Jonathan,

Thanks for the quick feedback. Note that you'll need to pull in the current
git head if you need --variant-input. The discordance at the end of the
contigs is concerning. If it looks like a regression let me know.

Erik
On May 21, 2015 8:20 AM, "jpuritz" notifications@github.com wrote:

Hey Erik,

I took a quick look at the fix and it seems to be working correctly. I'm
getting slightly different results from the two different versions ( on a
small test data set) in terms of genotypes and variant calls, but it is
somewhere over 99% concordance. The discordance is highest, it seems, at
sites that are INDEL or complex at the end of a reference contig. Not too
surprising. I'll keep looking into this next week and will keep you
informed.

Thanks,

Jon

—
Reply to this email directly or view it on GitHub
#163 (comment).

[github-actions]

This issue is marked stale because it has been open 120 days with no activity. Remove stale label or comment or this will be closed in 5 days

[github-actions]

This issue was closed for lack of activity. Feel free to re-open if someone feels like working on it.