Problems with failed contigs

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Problems with failed contigs

seoanezonjic
Hi
Currently i'm annotating a fish genome using the Maker 3 version and the execution don't finish properly. I execute Maker in a shared cluster trough MPI, asking for 96 cores (in order to not block the maker manager nor the file system) in 6 nodes of 16 cores each one. The work run 3 or 4 days and the output suddenly stops. The work remains 3 days more, but the index datastore is not modified and the only files that their attributes change are the lock files. When the job exceed the working time (7 days) it is cancelled. Then, I inspect the datastore index and count the uniq started and finished tags for each scaffold. The results are:
STARTED:3890
FINISHED:3378
So, there are about 500 scaffolds in which the analysis fails. When I inspect the STDERR maker output I see this error line repeated ~2000 times, at differents sites:
substr outside of string at /mnt/home/users/pab_001_uma/pedro/software/maker/bin/../lib/PhatHit_utils.pm line 850
and near this line, the following:
ERROR: Failed while annotating transcripts
My questions are, what can I do to annotate the failed scaffolds? and in addition, is the great amount of failed analysis the reason of the job freeze?
Thanks in advance





_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: Problems with failed contigs

seoanezonjic
Hi
I have tried to split the fasta file in small chunks and perform a execution for each chunk. Most of executions fail as I have described previously. I have inspect the results of a random chunk with 287 contigs and there are these error lines repeated 47 times:

substr outside of string at /mnt/home/users/pab_001_uma/pedro/software/maker/bin/../lib/PhatHit_utils.pm line 850.
--> rank=15, hostname=dx095
ERROR: Failed while annotating transcripts
ERROR: Chunk failed at level:1, tier_type:4
FAILED CONTIG:Sosen1_s1284
ERROR: Chunk failed at level:6, tier_type:0
FAILED CONTIG:Sosen1_s1284

Can you help me to fix this problem?
Thank you in advance
Pedro Seoane


De: p sz <[hidden email]>
Enviado: martes, 6 de marzo de 2018 9:30
Para: [hidden email]
Asunto: Problems with failed contigs
 
Hi
Currently i'm annotating a fish genome using the Maker 3 version and the execution don't finish properly. I execute Maker in a shared cluster trough MPI, asking for 96 cores (in order to not block the maker manager nor the file system) in 6 nodes of 16 cores each one. The work run 3 or 4 days and the output suddenly stops. The work remains 3 days more, but the index datastore is not modified and the only files that their attributes change are the lock files. When the job exceed the working time (7 days) it is cancelled. Then, I inspect the datastore index and count the uniq started and finished tags for each scaffold. The results are:
STARTED:3890
FINISHED:3378
So, there are about 500 scaffolds in which the analysis fails. When I inspect the STDERR maker output I see this error line repeated ~2000 times, at differents sites:
substr outside of string at /mnt/home/users/pab_001_uma/pedro/software/maker/bin/../lib/PhatHit_utils.pm line 850
and near this line, the following:
ERROR: Failed while annotating transcripts
My questions are, what can I do to annotate the failed scaffolds? and in addition, is the great amount of failed analysis the reason of the job freeze?
Thanks in advance





_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: Problems with failed contigs

Carson Holt-2
In reply to this post by seoanezonjic
First make sure you are using the most current version (3.01.02).

If you are providing a GFF3 issue, it may be an issue with one or more of the features you are giving. Finally, there is a possibility this is a BLAST report issue, as there is a BLAST truncation bug that appears to be fixed in NCBI BLAST 2.7.1+

—Carson

On Mar 6, 2018, at 2:30 AM, p sz <[hidden email]> wrote:

Hi
Currently i'm annotating a fish genome using the Maker 3 version and the execution don't finish properly. I execute Maker in a shared cluster trough MPI, asking for 96 cores (in order to not block the maker manager nor the file system) in 6 nodes of 16 cores each one. The work run 3 or 4 days and the output suddenly stops. The work remains 3 days more, but the index datastore is not modified and the only files that their attributes change are the lock files. When the job exceed the working time (7 days) it is cancelled. Then, I inspect the datastore index and count the uniq started and finished tags for each scaffold. The results are:
STARTED:3890
FINISHED:3378
So, there are about 500 scaffolds in which the analysis fails. When I inspect the STDERR maker output I see this error line repeated ~2000 times, at differents sites:
substr outside of string at /mnt/home/users/pab_001_uma/pedro/software/maker/bin/../lib/PhatHit_utils.pm line 850
and near this line, the following:
ERROR: Failed while annotating transcripts
My questions are, what can I do to annotate the failed scaffolds? and in addition, is the great amount of failed analysis the reason of the job freeze? 
Thanks in advance




_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org


_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org
Reply | Threaded
Open this post in threaded view
|

Re: Problems with failed contigs

seoanezonjic
I checked the maker version and it's the 3.01.02 that you suggest. I currently running maker for training SNAP (first round) so I'm not using gff files as input. Mi input files are all in fasta format, set in the genome, est and protein variables. In fact, one error line, saids that  the transcript analysis breaks the execution. I use the 2.2.30+ blast versión, maybe I have to use a newer version. Which version do you suggest me?
Thank you in advance

From: Carson Holt <[hidden email]>
Sent: Thursday, March 15, 2018 2:57:37 PM
To: p sz
Cc: [hidden email]
Subject: Re: [maker-devel] Problems with failed contigs
 
First make sure you are using the most current version (3.01.02).

If you are providing a GFF3 issue, it may be an issue with one or more of the features you are giving. Finally, there is a possibility this is a BLAST report issue, as there is a BLAST truncation bug that appears to be fixed in NCBI BLAST 2.7.1+

—Carson

On Mar 6, 2018, at 2:30 AM, p sz <[hidden email]> wrote:

Hi
Currently i'm annotating a fish genome using the Maker 3 version and the execution don't finish properly. I execute Maker in a shared cluster trough MPI, asking for 96 cores (in order to not block the maker manager nor the file system) in 6 nodes of 16 cores each one. The work run 3 or 4 days and the output suddenly stops. The work remains 3 days more, but the index datastore is not modified and the only files that their attributes change are the lock files. When the job exceed the working time (7 days) it is cancelled. Then, I inspect the datastore index and count the uniq started and finished tags for each scaffold. The results are:
STARTED:3890
FINISHED:3378
So, there are about 500 scaffolds in which the analysis fails. When I inspect the STDERR maker output I see this error line repeated ~2000 times, at differents sites:
substr outside of string at /mnt/home/users/pab_001_uma/pedro/software/maker/bin/../lib/PhatHit_utils.pm line 850
and near this line, the following:
ERROR: Failed while annotating transcripts
My questions are, what can I do to annotate the failed scaffolds? and in addition, is the great amount of failed analysis the reason of the job freeze? 
Thanks in advance




_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org


_______________________________________________
maker-devel mailing list
[hidden email]
http://box290.bluehost.com/mailman/listinfo/maker-devel_yandell-lab.org