Using Iso-seq data for Braker annotation

[Upendra19993]

Hi, I am trying to use iso-seq data for braker annotation. When I passed the job, within 30mins, I get an error.

I have attached the error here for your reference. Kindly let me know how to resolve this issue.

(base) [uqwwijes@bun113 braker3]$ singularity exec /scratch/project/qaafi-cnafs/braker3/braker3_lr.sif braker.pl --genome=M.indica_KP_genome.fasta.masked.fa --bam=KP_combined_SQANTI_uf_hq.bam --workingdir=KP_SQANTIBraker_sifrun3 --threads 24 --gff3 --softmasking --AUGUSTUS_CONFIG_PATH=/scratch/project/qaafi-cnafs/braker3/config # Tue Sep 24 06:45:24 2024: Log information is stored in file /scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/braker.log ERROR in file /opt/BRAKER/scripts/braker.pl at line 6291 Failed to execute: /opt/Augustus/bin//etraining --species=Sp_225 --AUGUSTUS_CONFIG_PATH=/scratch/project/qaafi-cnafs/braker3/config /scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/train.f.gb 1> /scratch/project_mnt/S0030/braker3/KP_SQANTIBraker_sifrun3/gbFilterEtraining.stdout 2>/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/errors/gbFilterEtraining.stderr****

The log file information also has attached. braker.log

Thank you very much!

Kind regards, Upendra.

[KatharinaHoff]

What is the content of these files?

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/gbFilterEtraining.stdout

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/errors/gbFilterEtraining.stderr

If there is a segmentation fault in the stderr file, please share (e.g. via e-mail):

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/train.f.gb

[Upendra19993]

Hi KatharinaHoff,

I have difficulties to attach the files you mentioned to this, but I have attached them and some other files saved in the 'error ' directory via an email.

Thank you very much!

Kind regards, Upendra.

[Upendra19993]

Hi Katharina Hoff,

Herewith, I have attached the gbFilterEtraining.stdout and train.f.gb files for your kind consideration.

The gbFilterEtraining.stderr you have mentioned is empty.

There are other files which have been saved under a directory names as 'error'. They all are empty.

Though the job ended without completing the whole process, there are some output files. Kindly let me know if it is good to have a look, I will share the files with you.

Thank you very much!

Best regards, Upendra. gbFilterEtraining.stdout https://drive.google.com/file/d/1GJdGiFxA67FZInPqlsbFwMDE-hICcrOS/view?usp=drive_web train.f.gb https://drive.google.com/file/d/1hZXHat9EWtbzsremIunvHJFboAe81hMi/view?usp=drive_web

On Tue, Sep 24, 2024 at 8:26 PM Katharina Hoff @.***> wrote:

What is the content of these files?

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/gbFilterEtraining.stdout

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/errors/gbFilterEtraining.stderr

If there is a segmentation fault in the stderr file, please share (e.g. via e-mail):

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/train.f.gb

— Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/BRAKER/issues/863#issuecomment-2370878395, or unsubscribe https://github.com/notifications/unsubscribe-auth/AWXPDELSWR2XHJAGD2SEJ43ZYE46PAVCNFSM6AAAAABOW74Q3WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZQHA3TQMZZGU . You are receiving this because you authored the thread.Message ID: @.***>

[Upendra19993]

gbFilterEtraining.stdout https://drive.google.com/file/d/1rl9g_eL_Y7xcX_HLkHumYc_FIzuWnVe0/view?usp=drive_web train.f.gb https://drive.google.com/file/d/1eGbcBiUFhxKujiy94DZM6gPY2r7uXJB2/view?usp=drive_web Hi Katharina Hoff,

I am sending the two files I attached to the previous email again for your reference. Let me know whether you have access to these files.

Thank you! Kind regards, Upendra.

On Wed, Sep 25, 2024 at 3:00 AM Upendra wijesundara < @.***> wrote:

Hi Katharina Hoff,

Herewith, I have attached the gbFilterEtraining.stdout and train.f.gb files for your kind consideration.

The gbFilterEtraining.stderr you have mentioned is empty.

There are other files which have been saved under a directory names as 'error'. They all are empty.

Though the job ended without completing the whole process, there are some output files. Kindly let me know if it is good to have a look, I will share the files with you.

Thank you very much!

Best regards, Upendra. gbFilterEtraining.stdout https://drive.google.com/file/d/1GJdGiFxA67FZInPqlsbFwMDE-hICcrOS/view?usp=drive_web train.f.gb https://drive.google.com/file/d/1hZXHat9EWtbzsremIunvHJFboAe81hMi/view?usp=drive_web

On Tue, Sep 24, 2024 at 8:26 PM Katharina Hoff @.***> wrote:

What is the content of these files?

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/gbFilterEtraining.stdout

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/errors/gbFilterEtraining.stderr

If there is a segmentation fault in the stderr file, please share (e.g. via e-mail):

/scratch/project_mnt/S0030/braker3/KP_SQANTI__Braker_sifrun3/train.f.gb

— Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/BRAKER/issues/863#issuecomment-2370878395, or unsubscribe https://github.com/notifications/unsubscribe-auth/AWXPDELSWR2XHJAGD2SEJ43ZYE46PAVCNFSM6AAAAABOW74Q3WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZQHA3TQMZZGU . You are receiving this because you authored the thread.Message ID: @.***>

[KatharinaHoff]

I have checked the train.f.gb file. It does not lead to etraining problems. Possibly, your species parameters are corrupt. If you could make them available, I could also have a look at them. I need the entire contents of the folder /scratch/project/qaafi-cnafs/braker3/config/species/Sp_225

[Upendra19993]

Hi Katharina,

I checked the folder /scratch/project/qaafi-cnafs/braker3/config/species/. But within the species folder, no folder is created for Sp_225. Kindly let me know how I can support you to find the error.

Thank you very much!

Best regards, Upendra.

On Sun, Sep 29, 2024 at 3:58 AM Katharina Hoff @.***> wrote:

I have checked the train.f.gb file. It does not lead to etraining problems. Possibly, your species parameters are corrupt. If you could make them available, I could also have a look at them. I need the entire contents of the folder /scratch/project/qaafi-cnafs/braker3/config/species/Sp_225

— Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/BRAKER/issues/863#issuecomment-2380848941, or unsubscribe https://github.com/notifications/unsubscribe-auth/AWXPDEOMY3TYW3OTNFW35BLZY3U3VAVCNFSM6AAAAABOW74Q3WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOBQHA2DQOJUGE . You are receiving this because you authored the thread.Message ID: @.***>

[KatharinaHoff]

Both the absence of the Sp_225 species folder and the error message in the etraining output file indicate that you are using an older version of an AUGUSTUS_CONFIG_PATH folder content. A json file is missing in that folder. This happens because some years ago, we switched to requiring that file, and older versions of augustus did not bring it.

If no AUGUSTUS_CONFIG_PATH is configured and provided to BRAKER (and if BRAKER is executed with the correct Augustus version from e.g. Singularity container), one of the first steps should be that BRAKER copies the content of the AUGUSTUS_CONFIG_PATH into ~/.augustus or into the current working directory.

Now, if you already have such an - older - folder sitting in home or in the working directory, please remove it so that BRAKER will know that it has to copy the folder contents.

Please use the Singularity container, ideally the most recent released version.

[Upendra19993]

Hi Katharina,

Thank you very much for your help in identifying the issue.

Just want to clarify some points,

1. First I need to check the Augustus version used for braker annotation, if it is an older version, I need to install the up to date version and have to give the path for the config folder and run it again
2. If I didn't specify the path for the config file, but have up to date Augustus version, then I need to delete if there are any other older versions of Augustus (config file folders)

Kindly advise.

Best regards, Upendra.

On Sun, Sep 29, 2024 at 10:48 PM Katharina Hoff @.***> wrote:

Both the absence of the Sp_225 species folder and the error message in the etraining output file indicate that you are using an older version of an AUGUSTUS_CONFIG_PATH folder content. A json file is missing in that folder. This happens because some years ago, we switched to requiring that file, and older versions of augustus did not bring it.

If no AUGUSTUS_CONFIG_PATH is configured and provided to BRAKER (and if BRAKER is executed with the correct Augustus version from e.g. Singularity container), one of the first steps should be that BRAKER copies the content of the AUGUSTUS_CONFIG_PATH into ~/.augustus or into the current working directory.

Now, if you already have such an - older - folder sitting in home or in the working directory, please remove it so that BRAKER will know that it has to copy the folder contents.

Please use the Singularity container.

— Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/BRAKER/issues/863#issuecomment-2381344367, or unsubscribe https://github.com/notifications/unsubscribe-auth/AWXPDEMEXEAP4LNAXQI4UJLZY7ZJHAVCNFSM6AAAAABOW74Q3WVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGOBRGM2DIMZWG4 . You are receiving this because you authored the thread.Message ID: @.***>

[tallnuttrbgv]

Deleting the config path prior to running does not work.

rm -rf /home/554/ta0341/.augustus/*

singularity exec -B ${PWD}:${PWD} /g/data/nm31/bin/braker3.sif braker.pl --gff3 --genome=/g/data/dy44/r21_genome_paralogs/genomes/hedy_masked.fasta --prot_seq=/g/data/dy44/r21_genome_paralogs/prothint/seed_proteins.faa --threads=48

Mon Oct 14 10:26:41 2024: WARNING: in file /opt/BRAKER/scripts/braker.pl at line 1933 AUGUSTUS_CONFIG_PATH/species (in this case /opt/Augustus/config//species) is not writeable. BRAKER will try to copy the AUGUSTUS config directory to a writeable location. chmod: cannot access '/home/554/ta0341/.augustus/species': No such file or directory ERROR in file /opt/BRAKER/scripts/braker.pl at line 2044 Failed to execute chmod u+w /home/554/ta0341/.augustus/species!

Tried other variations of copying configs which also fail:

https://github.com/Gaius-Augustus/BRAKER/issues/874

The directory structure that augustus is expecting is different from the one in the .sif.

[tallnuttrbgv]

Ok. Finally, this worked for me, just set the Augustus config path (in bash profile) to always be the current dir (so also works on PBS job node).

export AUGUSTUS_CONFIG_PATH=./