gnomon_training error

[guo-cheng]

Hi developers,

It's my test run:
python3 ui/egapx.py input_D_farinae_small.yaml -e singularity -w dfs_work -o dfs_out -lc local_cache -c ./egapx_config

ERROR ~ Error executing process > 'egapx:gnomon_plane:gnomon_training_iterations:gnomon_training_iteration:gnomon_training:run_gn

Caused by:
Process egapx:gnomon_plane:gnomon_training_iterations:gnomon_training_iteration:gnomon_training:run_gnomon_training terminate

Command executed:

mkdir -p output
lds2_indexer -source indexed -db ./indexed_lds
gnomon_training -b -asn -maxintron 100000 -nogenbank -lds2 ./indexed_lds -input gnomon_wnode.out -out output/hmm_params.asn

Command exit status:
1

Command output:
Models for training 0

Command error:
added loader: LDS2: indexed_lds (10)
Models for training 0
Not enough data for traiing 0

Work dir:
/public1/home/ligch/software/egapx/dfs_work/22/cfadd2b7a56aba4c19a8b6bb0dae9f

It's the failed dir:
dfs_work.tar.gz

gnomon_wnode.out:
https://drive.google.com/file/d/1z9l6hfXAuvzUI_vgaZwoi9XGmxsRDIYL/view?usp=drive_link

Thank you for your help!

Guo-Cheng

[pstrope]

Can you please try to re-download the files for local_cache, and then check how many files got downloaded?
python3 ui/egapx.py -dl -lc ../local_cache
and then
find ../local_cache -type f| wc -l

[guo-cheng]

Can you please try to re-download the files for local_cache, and then check how many files got downloaded? python3 ui/egapx.py -dl -lc ../local_cache and then find ../local_cache -type f| wc -l

Hi，
when I redownload the local_cache, the error:

Traceback (most recent call last):

File "/public1/home/ligch/software/egapx/ui/egapx.py", line 810, in
sys.exit(main(sys.argv))
File "/public1/home/ligch/software/egapx/ui/egapx.py", line 683, in main
download_egapx_ftp_data(args.local_cache)
File "/public1/home/ligch/software/egapx/ui/egapx.py", line 179, in download_egapx_ftp_data
ftpd.download_ftp_dir(FTP_EGAP_ROOT_PATH+f"/{line}", f"{local_cache_dir}/{line}")
File "/public1/home/ligch/software/egapx/ui/egapx.py", line 155, in download_ftp_dir
self.download_ftp_dir(next_ftp_name, next_local_name)
File "/public1/home/ligch/software/egapx/ui/egapx.py", line 158, in download_ftp_dir
r = self.download_ftp_file(next_ftp_name, next_local_name)
File "/public1/home/ligch/software/egapx/ui/egapx.py", line 82, in download_ftp_file
self.ftp.retrbinary("RETR {0}".format(ftp_name), f.write)
File "/public1/home/ligch/miniconda3/lib/python3.9/ftplib.py", line 438, in retrbinary
data = conn.recv(blocksize)
ConnectionResetError: [Errno 104] Connection reset by peer

[pstrope]

Can you check if you have access to the internet from your cluster?

[guo-cheng]

Can you please try to re-download the files for local_cache, and then check how many files got downloaded? python3 ui/egapx.py -dl -lc ../local_cache and then find ../local_cache -type f| wc -l

Hi,
Thank you for your reply. I also want to know how many files in local_cache normally. I have 868 files. Is it right?

[pstrope]

Yes, that sounds right. Now try running egapx.

[guo-cheng]

Yes, that sounds right. Now try running egapx.

Sorry, i still met this error. Could you please help me solve this problem?

ERROR ~ Error executing process > 'egapx:gnomon_plane:gnomon_training_iterations:gnomon_training_iteration:gnomon_training:run_gnomon_training'

Caused by:
Process egapx:gnomon_plane:gnomon_training_iterations:gnomon_training_iteration:gnomon_training:run_gnomon_training terminated with an error exit statu

Command executed:

mkdir -p output
lds2_indexer -source indexed -db ./indexed_lds
gnomon_training -b -asn -maxintron 100000 -nogenbank -lds2 ./indexed_lds -input gnomon_wnode.out -out output/hmm_params.asn

Command exit status:
1

Command output:
Models for training 0

Command error:
added loader: LDS2: indexed_lds (10)
Not enough data for traiing 0
Models for training 0

[boukn]

sorry for not replying sooner.

theres something weird about the gnomon output. there. Is your genome input exactly the one from the examples file, or has it been changed at all? Can you send the indexed/genomic.asn file that is just a symlink in the tar file?

[guo-cheng]

sorry for not replying sooner.

theres something weird about the gnomon output. there. Is your genome input exactly the one from the examples file, or has it been changed at all? Can you send the indexed/genomic.asn file that is just a symlink in the tar file?

GCF_020809275.1_ASM2080927v1_genomic.asn.zip

[boukn]

I found the problem, but its only really fixable in a code update. It will be fixed in the next release. Sorry about that

[guo-cheng]

I found the problem, but its only really fixable in a code update. It will be fixed in the next release. Sorry about that

Okay, I got it. I am looking forward the next release.

[Andy-B-123]

HI apologies for posting on a closed issue but I'm receiving this same error with the release of v0.3? 

I have tried to re-download the local cache with the new egapx.py file and can confirm I also have 886 files present. I saw in the release that there were some additional protein datasets for some vertebrates and plants, should the new local cache be a different size to the old one?

Strangely, I'm able to run the example dataset successfully with and without local cache, but I am unable to run a fish genome which I was able to annotate just last week with version 0.2!

A link to the contents of the egapx run including the input yaml file - https://github.com/Andy-B-123/datashare_repo/blob/main/contents.zip

I can look to share the genome + RNA-Seq too if that could help! The genome is a fish ~500MB fasta and the RNA-Seq is from 3x paired end illumina runs ~24GB fasta. The previous annotation (with version 0.2) was great, annotation BUSCOs ~95%, genome BUSCOs were ~95% and there was very little contamination according to OMark. 

Troublehsooting tips or steps appreciated! Thank you :-)

[boukn]

Can you send us the contents of /scratch3/bac050/temp_egapx_runs/5b/5e7828162ed9ae089310e92147024d
Including hidden .command.* files

[pstrope]

I can look to share the genome + RNA-Seq too if that could help!

@Andy-B-123 If you can put that up somewhere, we can test run that in our environment. Thanks!

[Andy-B-123]

all_dir_contents.zip

Attached a zip of the folder! Something that stands out to me is in .command.err:
added loader: LDS2: /scratch3/bac050/temp_egapx_runs/slurm_jobs/2260003/nxf.PEJOLiPNag/indexed_lds (10)
I checked and that file doesn't exist on my system? All that exists is the slurm directory with the job number.

Working on uploading my data to a shared area at the moment!

[guo-cheng]

HI apologies for posting on a closed issue but I'm receiving this same error with the release of v0.3? 

I have tried to re-download the local cache with the new egapx.py file and can confirm I also have 886 files present. I saw in the release that there were some additional protein datasets for some vertebrates and plants, should the new local cache be a different size to the old one?

Strangely, I'm able to run the example dataset successfully with and without local cache, but I am unable to run a fish genome which I was able to annotate just last week with version 0.2!

A link to the contents of the egapx run including the input yaml file - https://github.com/Andy-B-123/datashare_repo/blob/main/contents.zip

I can look to share the genome + RNA-Seq too if that could help! The genome is a fish ~500MB fasta and the RNA-Seq is from 3x paired end illumina runs ~24GB fasta. The previous annotation (with version 0.2) was great, annotation BUSCOs ~95%, genome BUSCOs were ~95% and there was very little contamination according to OMark. 

Troublehsooting tips or steps appreciated! Thank you :-)

Hi,
Because my internet is bad, I can't download the singularity image. Could you please give me the egapx-0.3.sif file?
Thanks a lot!

[Andy-B-123]

Apologies for the delay in uploading data, had some very frustrating internet issues at home. Back at work now and this should be the link to download the raw data for the run: https://filesender.aarnet.edu.au/?s=download&token=e70b0234-079f-41e2-b77d-a6da971ba129

Details:
- .yaml file is the .yaml file for the run, includes taxid ('303726')
- CSIRO-AGI_Nayr_v1.fasta.gz = reference genome for the fish
- Nayr_closest_proteins.fasta.gz = closest protein dataset for this fish 
- Na[1,2,3] are three sets of paired-end RNA-Seq data that I've tried to match the format required for the run.

The previous run with v0.2 seemed like a really nice annotation from the stats and overall profile, just strange that I'm hitting this Gnomon error now! I am running this with local cache, if it runs on your system then I'll keep troubleshooting on my end :-)

Also wanted to mention I'd be happy to test this genome for GenBank submission? I'm working as part of the CSIRO AGI group here in Australia and we're going to be doing a lot of genome + annotation uploads!

[boukn]

Sorry, I just fixed this internally yesterday. We are going to release a 0.3.1 shortly that includes a fix for this.

[etvedte]

I'd be happy to test this genome for GenBank submission?

@Andy-B-123 can you send your contact info to eric.tvedte@nih.gov?