Error while trying to run braker2 with --UTR=on

[aswathyseb]

Hi,

I am running the example data with the following command.

braker.pl --genome genome.fa --prot_seq proteins.fa --prg gth --bam RNAseq.bam --gth2traingenes --softmasking --cores 22 --UTR=on --gff3 --workingdir=braker_with_train_UTR

And I get the following error. ERROR in file /home/aswathy/src/BRAKER/scripts/braker.pl at line 9115 Could not open file /my/path/to/example/braker_with_train_UTR/Ppri5_utr.job.lst!

As per the manual I have obtained and running Augustus from github and it is version 3.4.0

[tomasbruna]

Hi,

did you also update BRAKER2 from its master branch here on GitHub? We are working on a new BRAKER2 release, but for now, the most updated BRAKER2 version needs to be obtained from the master.

The issue could also be caused by using GenomeThreader for proteins (--prg gth). This mode of protein alignments is not actively supported anymore, it is possible that it's not properly working in combination with UTRs.

The safest way of adding UTRs to your predictions (using both proteins and RNA-Seq) is described here https://github.com/Gaius-Augustus/BRAKER#testing-braker-with-proteins-of-any-evolutionary-distance-and-rna-seq. Unless you really need to use GenomeThreader, I recommend you try that.

Best regards, Tomas

[aswathyseb]

Thank you for the quick response.

Yes, I obtained both Augustus and Braker2 from github.

I see that to run braker2 with proteins of any evolutionary distance pipeline and to avoid using --prg gth , protein sequences from multiple species are needed. I want to use proteins from just one related species along with rnaseq data from the same species. Can I do that with 'proteins of any evolutionary distance pipeline' ?

[KatharinaHoff]

You can concatenate the proteins of you close relative to an ortho db partition and then use prothint. Higher accuracy expected than with only that one protein set.

On Wed 16. Dec 2020 at 16:43, Aswathy notifications@github.com wrote:

Thank you for the quick response.

Yes, I obtained both Augustus and Braker2 from github.

I see that to run braker2 with proteins of any evolutionary distance pipeline and to avoid using --prg gth , protein sequences from multiple species are needed. I want to use proteins from just one related species along with rnaseq data from the same species. Can I do that with 'proteins of any evolutionary distance pipeline' ?

— 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/310#issuecomment-746501340, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJMC6JB46A4UY5WIJDQVEGLSVDIRVANCNFSM4U5Z67KQ .

[tomasbruna]

Just to expand on Katharina's comment, here are instructions on how to prepare the OrthoDB protein set: https://github.com/gatech-genemark/ProtHint#protein-database-preparation.

Using proteins from just one related species would also work with the 'proteins of any evolutionary distance pipeline', but as Katharina said, adding more species to the protein set generally improves the prediction accuracy.

[aswathyseb]

Ok, thanks I will try that.

But I ran the data given in the example with ProtHint and --UTR=on with the command braker.pl --genome genome.fa --prot_seq proteins.fa --bam RNAseq.bam --etpmode --softmasking --cores 22 --UTR=on --gff3 --workingdir=braker_default_with_UTR

and I get the same error. ERROR in file /home/aswathy/src/BRAKER/scripts/braker.pl at line 9115 Could not open file /path/to/example/braker_default_with_UTR/Ppri5_utr.job.lst!

So it looks like regardless of the way braker2 is running, --UTR=on is giving this error.

[tomasbruna]

I see, the --UTR=on is still an experimental feature, we will address this bug in the future, thanks for reporting it.

The mode of adding UTRs to existing BRAKER2 predictions (--addUTR=on) should be more stable, I hope it will work for you.

Best, Tomas

[aswathyseb]

OKay, thank you.

[ricardo-aaron]

this sounds like (closed) issue #152

[KatharinaHoff]

It is not closed. Fixing is postponed.

On Sat 29. May 2021 at 04:59, Ricardo A. Chávez Montes < @.***> wrote:

this sounds like (closed) issue #152 https://github.com/Gaius-Augustus/BRAKER/issues/152

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/Gaius-Augustus/BRAKER/issues/310#issuecomment-850757969, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJMC6JCWLVMPBU6RJZ6BGF3TQBKBTANCNFSM4U5Z67KQ .