esource: invalid source key: RM

[francicco]

Hi,

I'm many genomes to annotate with BRAKER only few fail. The error I get from one of the is in the Third etraining - now with CRF step. Etraining throw me this error:

/mnt/storage/home/tk19812/scratch/software/Augustus/bin/etraining: ERROR FeatureCollection::esource: invalid source key: RM

Anything I can do?

Thanks F

[KatharinaHoff]

Can you please provide the braker.log file and if there is a etraining.stdout & etraining.stderr file, these files, too?

It seems a weird error because the hints should not go into training at all. RM is our source key for repeat masker information. It's a hints source key.

On Tue, Feb 23, 2021 at 9:44 AM Francesco Cicconardi < [email protected]> wrote:

Hi,

I'm many genomes to annotate with BRAKER only few fail. The error I get from one of the is in the Third etraining - now with CRF step. Etraining throw me this error:

/mnt/storage/home/tk19812/scratch/software/Augustus/bin/etraining: ERROR FeatureCollection::esource: invalid source key: RM

Anything I can do?

Thanks F

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/BRAKER/issues/325, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJMC6JE2KK7DXRZEAWGRXYTTANTFXANCNFSM4YCB47ZQ .

[francicco]

Hi @KatharinaHoff

Here are the files:

braker.log crftraining.stderr.txt crftraining.stdout.txt

Thanks F

[francicco]

Hi Katharina,

Any update?

Cheers F

[monpau]

Hi, I'm also experiencing the same error. My braker.log and crftraining.std* look exactly the same as Francisco's.

Any suggestions?

Thanks

[tiramisutes]

Hi, I'm also experiencing the same error.

[WenChangYu]

Hi, When I use --crf option, I encounter the same error.

Any suggestions?

[kmkocot]

I am having the same issue when trying to use --crf here. I'm using braker.pl version 2.1.6 and Augustus 3.4.0.

[amvarani]

same problem here

[KatharinaHoff]

@MarioStanke is it possible that AUGUSTUS treats softmasking differently during GHMM and CRF training?

[kmkocot]

Hi all. Switching back from Augustus 3.4.0 to 3.3.3 fixed this issue for me for what it's worth.

[RenscoHogers]

Hi all, indeed switching to 3.3.3 worked to fix this issue, but is there any progress on fixing this issue in 3.4.0?