Open jolespin opened 4 years ago
JaninaNii
commented
4 years ago I'm also getting a segmentation fault error when running the following command. Shorter predictions (predictionStart=1000001, predictionEnd=1100000) work nicely. I checked that CPU% goes to 100% before getting this error. I'm using Augustus directly on linux server.
augustus --species=candida_albicans --predictionStart=1000001 --predictionEnd=1200000 CP017623.fasta > augustus.abinitio.gff
KatharinaHoff
commented
4 years ago @MarioStanke please have a look.
jolespin
commented
4 years ago Could this be a Python 2/3 error?
What could be going on here?
INFO [augustus] /bin/sh: line 1: 5413 Segmentation fault (core dumped) augustus --stopCodonExcludedFromCDS=False --codingseq=1 --proteinprofile=/usr/local/scratch/METAGENOMICS/jespinoz/db/funannotate_db/dikarya/prfl/EOG09261JCQ.prfl --predictionStart=306979 --predictionEnd=317740 --species=anidulans './tmp/CP022972.1saccharomyces_664759016_.temp' > /local/ifs2_projdata/0568/projects/PLANKTON/illumina_aallen/jespinoz/testing/test-predict_239889/annotate/predict_misc/busco/run_saccharomyces/augustus_output/predicted_genes/EOG09261JCQ.out.1 2>> /local/ifs2_projdata/0568/projects/PLANKTON/illumina_aallen/jespinoz/testing/test-predict_239889/annotate/predict_misc/busco/run_saccharomyces/augustus_output/augustus.logIt seems likely to be a compilation error from this discussion. https://github.com/nextgenusfs/funannotate/issues/312
KatharinaHoff
commented
4 years ago No.
On Thu 22. Aug 2019 at 23:22, Josh L. Espinoza notifications@github.com wrote:
Could this be a Python 2/3 error?
— You are receiving this because you commented.
Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/Augustus/issues/75?email_source=notifications&email_token=AJMC6JDDPLRG73KEGD4RQLTQF37RTA5CNFSM4IOR4IWKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD46N6CA#issuecomment-524082952, or mute the thread https://github.com/notifications/unsubscribe-auth/AJMC6JEBJHD4E7HLJIIZMWLQF37RTANCNFSM4IOR4IWA .
jolespin
commented
4 years ago Is there a test I can run to make sure my installation is working properly?
On Aug 22, 2019, at 8:44 PM, Katharina Hoff notifications@github.com wrote:
No.
On Thu 22. Aug 2019 at 23:22, Josh L. Espinoza notifications@github.com wrote:
Could this be a Python 2/3 error?
— You are receiving this because you commented.
Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/Augustus/issues/75?email_source=notifications&email_token=AJMC6JDDPLRG73KEGD4RQLTQF37RTA5CNFSM4IOR4IWKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD46N6CA#issuecomment-524082952, or mute the thread https://github.com/notifications/unsubscribe-auth/AJMC6JEBJHD4E7HLJIIZMWLQF37RTANCNFSM4IOR4IWA .
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.
jolespin
commented
4 years ago I've opened a question on on biostars https://www.biostars.org/p/395618/#395947 but they closed the question redirecting me this GitHub post. I'm sort of in a loop trying to figure this out. Any help would be greatly appreciated.
KatharinaHoff
commented
4 years ago An AUGUSTUS segmentation fault is typically caused by using a binary that was compiled on some system on another system — or it is a real bug caused by some particular input data. In order to reproduce such a bug, we will need a toy data set that shows the problem on one of our own machines (including the species parameter set). In any case, it can take weeks or months until someone has time to look into such a bug.
So my first suggestion is that you clone AUGUSTUS from github and compile on your own machine, check whether the problem persists. If yes, please provide toy data and eventually, someone might look into it.
Katharina
On Mon 26. Aug 2019 at 22:29, Josh L. Espinoza notifications@github.com wrote:
I've opened a question on on biostars https://www.biostars.org/p/395618/#395947 but they closed the question redirecting me this GitHub post. I'm sort of in a loop trying to figure this out. Any help would be greatly appreciated.
— You are receiving this because you commented.
Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/Augustus/issues/75?email_source=notifications&email_token=AJMC6JDWUEMCHVUGEZOD5T3QGQ4J5A5CNFSM4IOR4IWKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5FSQBA#issuecomment-525019140, or mute the thread https://github.com/notifications/unsubscribe-auth/AJMC6JEENSP5I5TGO5PUQALQGQ4J5ANCNFSM4IOR4IWA .
jolespin
commented
4 years ago Sounds good, I used a conda installation but I can try installing it separately and then throwing all of the necessary files into $CONDA_PREFIX/envs/funannotate_env/bin
Which files/folders are required to be added into my path aside from AUGUSTUS_CONFIG_PATH to properly function?
KatharinaHoff
commented
4 years ago The path to the AUGUSTUS binary must be in your $PATH variable (Augustus/bin). It is probably not related to your issue, but you might want to add Augustus/scripts, too. Make sure no old install is still in your PATH.
On Mon 26. Aug 2019 at 23:07, Josh L. Espinoza notifications@github.com wrote:
Sounds good, I used a conda installation but I can try installing it separately and then throwing all of the necessary files into $CONDA_PREFIX/envs/funannotate_env/bin
Which files/folders are required to be added into my path aside from AUGUSTUS_CONFIG_PATH to properly function?
— You are receiving this because you commented.
Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/Augustus/issues/75?email_source=notifications&email_token=AJMC6JGLOAXOE76TEIC4BCTQGRAXJA5CNFSM4IOR4IWKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5FV23A#issuecomment-525032812, or mute the thread https://github.com/notifications/unsubscribe-auth/AJMC6JC3Q5BUZMHXKQGSCX3QGRAXJANCNFSM4IOR4IWA .
jolespin
commented
4 years ago Awesome, thanks for you help. Is there a test command I could run or a toy dataset that should run easily? Just so I can rule out all of the options during debugging my environment.
KatharinaHoff
commented
4 years ago You could use the example genome file from BRAKER. Should work with fly parameters ab initio... or make it smaller... or predict in a small range only
On Mon 26. Aug 2019 at 23:32, Josh L. Espinoza notifications@github.com wrote:
Awesome, thanks for you help. Is there a test command I could run or a toy dataset that should run easily? Just so I can rule out all of the options during debugging my environment.
— You are receiving this because you commented.
Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/Augustus/issues/75?email_source=notifications&email_token=AJMC6JACXM5M62HFCAFPGZDQGRDVXA5CNFSM4IOR4IWKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD5FX7FY#issuecomment-525041559, or mute the thread https://github.com/notifications/unsubscribe-auth/AJMC6JA4Z46FRITJJK2ZCILQGRDVXANCNFSM4IOR4IWA .
I'm using the
funannotatepipeline and it fails during theaugustusstep. I found which command it was running and tried running it outside offunannotateand got the following error. How do I fix this? I installed on linux server via conda.