Failing during optimize_augustus.pl

[weedcentipede]

Hey, it's me again,
I tried to generate a model for my species and during the optimize_augustus.pl, I got the following error:

bio@biocomp04:~/Documents/Purpureocillium/March2019$ perl /home/bio/apps/augustus-3.3.2/scripts/optimize_augustus.pl --rounds=5 --species=Purp --kfold=8 --AUGUSTUS_CONFIG_PATH=/home/bio/apps/augustus-3.3.2/config --onlytrain=/home/bio/Documents/Purpureocillium/March2019/braker/Purp/train.gb.train.train --cpus=8 /home/bio/Documents/Purpureocillium/March2019/braker/Purp/train.gb.train.test
Splitting training file into 8 buckets...
Reading in the meta parameters used for optimization from /home/bio/apps/augustus-3.3.2/config/species/generic/generic_metapars.cfg...
Reading in the starting meta parameters from /home/bio/apps/augustus-3.3.2/config/species/Purp/Purp_parameters.cfg...
bucket Segmentation fault (core dumped)
2 Segmentation fault (core dumped)
Segmentation fault (core dumped)
3 Segmentation fault (core dumped)
8 5 Segmentation fault (core dumped)
1 Segmentation fault (core dumped)
6 Segmentation fault (core dumped)
Segmentation fault (core dumped)
4 7 Could not read the accuracy values out of predictions.txt when processing bucket 1. at /home/bio/apps/augustus-3.3.2/scripts/optimize_augustus.pl line 1224.

Later I used the following command, and Braker2 executed perfectly,

bio@biocomp04:~/Documents/Purpureocillium/March2019$ ~/apps/BRAKER-2.1.2/scripts/braker.pl --genome=./working-genomes/Purp-polished-abyss.fasta.masked --bam=./Purp-annotation/Purp-RNAseq-to-genome-unmasked.bam --species=Purp --cores=4 --AUGUSTUS_CONFIG_PATH=/home/bio/apps/augustus-3.3.2/config --GENEMARK_PATH=/home/bio/apps/gm_et_linux_64/gmes_petap --fungus --skipOptimize --useexisting

What do you think it could be,
Am I executing it wrong?
Cheers,
Luis Alfonso.

[KatharinaHoff]

Hi Alfonso, we have seen optimize_augustus.pl fail in two cases: first, when the AUGUSTUS binary cannot be located ( #17), second, when the number of training genes is very large ( Gaius-Augustus/Augustus#42). The second command that you can ran skips optimize_augustus.pl ... so it is not surprising that it worked... be aware, parameters for your species have never been optimized (unless you did that outside of BRAKER). Thanks, Katharina

…

On Tue, Apr 2, 2019 at 3:55 AM Luis A. Arteaga-Figueroa < ***@***.***> wrote: Hey, it's me again, I tried to generate a model for my species and during the optimize_augustus.pl, I got the following error: ***@***.***:~/Documents/Purpureocillium/March2019$ perl /home/bio/apps/augustus-3.3.2/scripts/optimize_augustus.pl --rounds=5 --species=Purp --kfold=8 --AUGUSTUS_CONFIG_PATH=/home/bio/apps/augustus-3.3.2/config --onlytrain=/home/bio/Documents/Purpureocillium/March2019/braker/Purp/train.gb.train.train --cpus=8 /home/bio/Documents/Purpureocillium/March2019/braker/Purp/train.gb.train.test Splitting training file into 8 buckets... Reading in the meta parameters used for optimization from /home/bio/apps/augustus-3.3.2/config/species/generic/generic_metapars.cfg... Reading in the starting meta parameters from /home/bio/apps/augustus-3.3.2/config/species/Purp/Purp_parameters.cfg... bucket Segmentation fault (core dumped) 2 Segmentation fault (core dumped) Segmentation fault (core dumped) 3 Segmentation fault (core dumped) 8 5 Segmentation fault (core dumped) 1 Segmentation fault (core dumped) 6 Segmentation fault (core dumped) Segmentation fault (core dumped) 4 7 Could not read the accuracy values out of predictions.txt when processing bucket 1. at /home/bio/apps/augustus-3.3.2/scripts/optimize_augustus.pl line 1224. Later I used the following command, and Braker2 executed perfectly, ***@***.***:~/Documents/Purpureocillium/March2019$ ~/apps/BRAKER-2.1.2/scripts/braker.pl --genome=./working-genomes/Purp-polished-abyss.fasta.masked --bam=./Purp-annotation/Purp-RNAseq-to-genome-unmasked.bam --species=Purp --cores=4 --AUGUSTUS_CONFIG_PATH=/home/bio/apps/augustus-3.3.2/config --GENEMARK_PATH=/home/bio/apps/gm_et_linux_64/gmes_petap --fungus --skipOptimize --useexisting What do you think it could be, Am I executing it wrong? Cheers, Luis Alfonso. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#30>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AlgvJGCUYqvKbGDsDfeLTi5Qx1XeGTeFks5vcriigaJpZM4cW-ID> .

[weedcentipede]

Hi Katarina,
Apparently the issue raised because I had two versions of augustus and one of was not suitable for the analysis,

Cheers,
Luis Alfonso.