[ensembl-dev] Targetted Genewise

Alex Humberto Di Genova Bravo digenova at gmail.com
Thu Oct 28 20:43:34 BST 2010


Dear Ensembl Developers:

I am working in a local installataion of the ensembl pipeline. I  
complete the raw compute stage of the pipeline without problem and now
I trying to run the Targetted stage of the gene build process, This is  
my actual configuration of this analisys:

      TargettedGenewise => {

              PAF_LOGICNAMES => ['BestPmatch'],
              OUTPUT_BIOTYPE => 'targetted',
              GENEWISE_PARAMETERS => {
                                      # pass parameters go genewise  
here, i.e. -program =>"/usr/local/ensembl/bin/genewiseXXX"
                                      # for more options which can be  
passed see Runnable/Genewise.pm
                                      #-program=>'/genoma/gendb/ 
ensembl/bin/genewise'
                                      #-endbias => 1,
                                      -gene => '/genoma/gendb/ensembl/ 
lib/cfgwise/human.gf',
                                      -matrix => '/genoma/gendb/ 
ensembl/lib/cfgwise/BLOSUM80.bla',
                                      -codon => '/genoma/gendb/ensembl/ 
lib/cfgwise/codon.table',
                                      -gap => 20,
                                      -extension => 8,
                                      -splice_model => 1
                                     },
              MINIGENEWISE_PARAMETERS => {
                                          -terminal_padding => 20000,
                                          -exon_padding => 200,
                                          -minimum_intron => 1000,
                                         },
              MULTIMINIGENEWISE_PARAMETERS =>{
                                               -minimum_feature_length  
=> 50,
                                             },
              FILTER_PARAMS => {
                                -max_exon_length => '20000',
                                -multi_exon_min_coverage => '25',
                                -single_exon_min_coverage => '80',
                                -max_intron_length => '200000',
                                -min_split_coverage => 95,
                                -max_low_complexity => 101,
                               },
              SEQFETCHER_PARAMS => {
                                    -db => ['/genoma/gendb/ensembl/ 
testgenome/protein_danio_rerio_index']
                                   },
              BLASTMINIGENEWISE_PARAMETERS => {},
              LIMIT_TO_FEATURE_RANGE => 1,
              FEATURE_RANGE_PADDING => 20000,
             },

Stored in the Config file of BlasMiniGenewise Module, The binaries  
(SEQFETCHERT , xdformat, wutblastn and  genewise) run in my test  
genome without
problems, I confirm that comenting the lines on the  
runnables(Blast.pm, genewise.pm and BlastMiniGenewise.pm) that delete  
the temporal files.
Currently my databases configuration comprise  the REFERENCE_DB and a  
separate GENEWISE_DB as is specified in the gene build readme or in
databases readme.

I genereta the input IDs for TargettedGenewise on slice level of 1MB  
using the make_input_id script, So When I check the job error there is  
a lot of warnings that end whit an error of undefined seq:


-------------------- WARNING ----------------------
MSG: Transcript has no translation, or no start exon - maybe a  
pseudogene?
FILE: Tools/GeneBuildUtils/TranscriptUtils.pm LINE: 1921
CALLED BY: Analysis/RunnableDB/BlastMiniGenewise.pm  LINE: 648
Ensembl API version = 60
---------------------------------------------------
1
-------------------- WARNING ----------------------
MSG: Transcript has no translation, or no start exon - maybe a  
pseudogene?
FILE: Tools/GeneBuildUtils/TranscriptUtils.pm LINE: 1921
CALLED BY: Analysis/RunnableDB/BlastMiniGenewise.pm  LINE: 648
Ensembl API version = 60
---------------------------------------------------
Non-coding transcript does not have -1 phases at /genoma/gendb/ensembl/ 
ensembl-analysis/modules/Bio/EnsEMBL/Analysis/Tools/GeneBuildUtils/ 
TranscriptUtils.pm line 466.Can't call method "seq" on an undefined  
value at /genoma/gendb/ensembl/ensembl-analysis/modules/Bio/EnsEMBL/ 
Analysis/Tools/GeneBuildUtils/TranslationUtils.pm line 201.

RUNNING: Lost the will to live Error
Job 16910 failed: [
-------------------- EXCEPTION --------------------
MSG: Problems running BlastMiniGenewise for  
chromosome:NCBI36:1:25637249:26623296:1 [Can't call method "seq" on an  
undefined value at /genoma/gendb/ensembl/ensembl-analysis/modules/Bio/ 
EnsEMBL/Analysis/Tools/GeneBuildUtils/Tr
anslationUtils.pm line 201.
]

STACK Bio::EnsEMBL::Pipeline::Job::run_module /genoma/gendb/ensembl/ 
ensembl-pipeline/modules/Bio/EnsEMBL/Pipeline/Job.pm:634
STACK (eval) /genoma/gendb/ensembl/ensembl-pipeline/modules/Bio/ 
EnsEMBL/Pipeline/runner.pl:219
STACK main::run_jobs_with_lsfcopy /genoma/gendb/ensembl/ensembl- 
pipeline/modules/Bio/EnsEMBL/Pipeline/runner.pl:218
STACK toplevel /genoma/gendb/ensembl/ensembl-pipeline/modules/Bio/ 
EnsEMBL/Pipeline/runner.pl:128
Ensembl API version = 60
---------------------------------------------------
]Finished job 16910


and there is no feature stored in  GENEWISE_DB,  this happend for all  
of my jobs, Any suggestion or advice will be very usefull
for me.


Thank in advance

Alex Di Génova.
Ingeniero en Bioinformatica.
Laboratorio de Bioinformatica y Matemática del Genoma.
Centro de Modelamiento Matemático.
Facultad de Ciencias Físicas y Matemáticas, Universidad de Chile.
Avenida Blanco Encalada 2120, Piso 6.





More information about the Dev mailing list