Augustus icon indicating copy to clipboard operation
Augustus copied to clipboard

augustus: ERROR Properties::getProperty(): cannot convert value "#" of key "/Constant/opalprob" into double

Open bioinformaticspcj opened this issue 4 years ago • 3 comments

Dear authors: I am using autoAugTrain.pl to train a new model for a reference genome, but I met an error: "augustus: ERROR Properties::getProperty(): cannot convert value "#" of key "/Constant/opalprob" into double" The command is as follows: "augustus-3.3.1/scripts/autoAugTrain.pl --trainingset=GCF_009769535.1_rThaEle1.pri_genomic.gff --genome=GCF_009769535.1_rThaEle1.pri_genomic.fna --workingdir=Tele --useexisting --species=garter_snake" I do not know why. Could you help me. Thanks a lot.

Looking forward to your timely reply. Best

bioinformaticspcj avatar Jun 09 '20 07:06 bioinformaticspcj

Hello,

I am facing the same issue! (Properties::getProperty(): cannot convert value "#" of key "/Constant/opalprob" into double.)

As i am working with non-regular fungi, i am trying to train augustus with the following command :

autoAug.pl --species=trichophyton_benhamiae --genome=simplify.Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.dna.toplevel.fa --cdna=simplify.Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.cdna.all.fa --trainingset=Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.47.nonchromosomal.dat

and I have the following output :

1 Checking fasta headers in file /sandbox/users/bbidon/trichophyton_benhamiae/augustus/simplify.Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.dna.toplevel.fa... 1 Checking fasta headers in file /sandbox/users/bbidon/trichophyton_benhamiae/augustus/simplify.Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.cdna.all.fa... 2 All necessary directories have been created unter /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/trainingSet. 1 Aligning cDNA to genome with BLAT... 2 Loaded 22224160 letters in 68 sequences 2 Searched 11833865 bases in 8000 sequences WARNING: Could not successfully find and run pslCDnaFilter. Please install this program. Will continue anyways with unfiltered alignments. Expect worse results. 1 Creating hints from cDNA alignments.... 2 Running perl /sandbox/users/bbidon/.conda/envs/genomic/bin/blat2hints.pl --in=../cdna/cdna.f.psl --out=hints.E.gff --minintronlen=35 --trunkSS 1>blat2hints.stdout 2>blat2hints.stderr ... Finished!

1 ####### Step 1: Training AUGUSTUS (no UTR models) #######

2 perl /sandbox/users/bbidon/.conda/envs/genomic/bin/autoAugTrain.pl -t=/sandbox/users/bbidon/trichophyton_benhamiae/augustus/Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.47.nonchromosomal.dat -s=trichophyton_benhamiae -g=/sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/seq/genome_clean.fa -w=/sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug -v -v --opt=1

/sandbox/users/bbidon/trichophyton_benhamiae/augustus/Trichophyton_benhamiae_cbs_112371_gca_000151125.ASM15112v2.47.nonchromosomal.dat appears to be in corrupt Genbank format. 'ORIGIN' missing 2 The input training set file has Genbank format. 2 Running command "perl /sandbox/users/bbidon/.conda/envs/genomic/bin/new_species.pl --species=trichophyton_benhamiae --silent 1>new_species.stdout" to create AUGUSTUS config files for new species trichophyton_benhamiae 2 Will create parameters for a EUKARYOTIC species! 2 The necessary files for training trichophyton_benhamiae have been created. 2 Now training AUGUSTUS parameters for trichophyton_benhamiae. 1 training.gb contains 68 sequences and 7980 genes, each sequence contains 117.352941176471 gene(s) on average. 2 Creating training.gb.train and training.gb.test with "perl /sandbox/users/bbidon/.conda/envs/genomic/bin/randomSplit.pl training.gb 1"... 2 Created training.gb.train and training.gb.test under /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain. 1 test/evaluation set training.gb.test contains 1 sequences and 1 genes. 1 training set training.gb.train contains 67 sequences and 7979 genes. 2 Creating training.gb.train.test and training.gb.onlytrain: 1 randomly selecting 2 genes from the training set training.gb.train...2 training.gb.train.test and training.gb.onlytrain have been made under /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain. 1 training.gb.train.test contains 2 sequences and 2 genes. 1 training.gb.onlytrain contains 65 sequences and 65 genes. 2 Seting value of "stopCodonExcludedFromCDS" in trichophyton_benhamiae_parameters.cfg to "true" 2 First try with etraining: etraining --species=trichophyton_benhamiae training.gb.train >train.out ... Finished! 2 Setting frequency of stop codons to tag=, taa=, tga=. 2 First evaluation of parameters ... 2 Excuting "augustus --species=trichophyton_benhamiae /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain/training/training.gb.test > /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain/training/test/augustus.1.out" ... augustus: ERROR Properties::getProperty(): cannot convert value "#" of key "/Constant/opalprob" into double.

bidonb avatar Jun 18 '20 14:06 bidonb

failed to execute: augustus --species=trichophyton_benhamiae /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain/training/training.gb.test > /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain/training/test/augustus.1.out! failed to execute: No such file or directory

P.S. I don't understand. It seems that genbank file is unccorect. I have taken it from ensembl for Trichophyton benhamiae, so I don't know wath to do next.

Do you have idees to help me. Thank a lot. Baptiste

bidonb avatar Jun 18 '20 14:06 bidonb

You cannot train AUGUSTUS on a genbank file retrieved from public data repositories such as GenBank. If you want to do something like this, you have to provide genome & gtf file. AUGUSTUS has expectations to a GenBank file that are not satisfied by the repository genbank format.

On Thu, Jun 18, 2020 at 4:55 PM bidonb [email protected] wrote:

failed to execute: augustus --species=trichophyton_benhamiae /sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain/training/training.gb.test

/sandbox/users/bbidon/trichophyton_benhamiae/augustus/autoAug/autoAugTrain/training/test/augustus.1.out! failed to execute: No such file or directory

P.S. I don't understand. It seems that genbank file is unccorect. I have taken it from ensembl for Trichophyton benhamiae, so I don't know wath to do next.

Do you have idees to help me. Thank a lot. Baptiste

— 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/Augustus/issues/155#issuecomment-646073095, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJMC6JDK2QHWXQZN3UUB4BDRXITF3ANCNFSM4NZEGHJQ .

KatharinaHoff avatar Jun 19 '20 09:06 KatharinaHoff