KOBAS on the Command Line

Getting the databases

To run the tool you need some public data. The files can be downloaded directly from the KOBAS homepage. These directories are also available as two tar archives in the CyVerse Data Store. The files are best downloaded with iCommands. Once iCommands is setup you can use ‘iget’ to download the data.

  1. seq_pep.tar: species-specific BLAST databases used by KOBAS
iget /iplant/home/shared/iplantcollaborative/protein_blast_dbs/kobas/seq_pep.tar

tar -xf seq_pep.tar
  1. sqlite3.tar: species-specific annotation databases used by KOBAS
iget /iplant/home/shared/iplantcollaborative/protein_blast_dbs/kobas/sqlite3.tar

tar -xf sqlite3.tar

Note

The above commands should result in two directories (seq_pep and sqlite3) each containing many files. There is no need to unzip the .gz files.

Container Technologies

KOBAS is provided as a Docker container.

A container is a standard unit of software that packages up code and all its dependencies so the application runs quickly and reliably from one computing environment to another.

There are two major containerization technologies: Docker and Singularity.

Docker containers can be run with either technology.

Running KOBAS using Docker

About Docker

  • Docker must be installed on the computer you wish to use for your analysis.
  • To run Docker you must have ‘root’ permissions (or use sudo).
  • Docker will run all containers as ‘root’. This makes Docker incompatible with HPC systems (see Singularity below).
  • Docker can be run on your local computer, a server, a cloud virtual machine (such as CyVerse Atmosphere) etc. Docker can be installed quickly on an Atmosphere instance by typing ‘ezd’.
  • For more information on installing Docker on other systems see this tutorial: Installing Docker on your machine.

Getting the KOBAS container

The KOBAS tool is available as a Docker container on Docker Hub: KOBAS container

The container can be pulled with this command:

docker pull agbase/kobas:3.0.3_0

Remember

You must have root permissions or use sudo, like so:

sudo docker pull agbase/kobas:3.0.3_0

Running KOBAS with Data

Getting the Help and Usage Statement

sudo docker run --rm -v $(pwd):/work-dir agbase/kobas:3.0.3_0 -h

See Help and Usage Statement

Tip

There are 3 directories built into this container. These directories should be used to mount data.
  • /work-dir
  • /seq_pep
  • /sqlite3

KOBAS can perform two tasks - annotate (-a) - identify (enrichment) (-g)

KOBAS can also run both task with a single command (-j).

Annotate Example Command

sudo docker run \
--rm \
-v /home/amcooksey/i5k/seq_pep:/seq_pep \
-v /home/amcooksey/i5k/sqlite3:/sqlite3 \
-v $(pwd):/work-dir \
agbase/kobas:3.0.3_0 \
-a
-i AROS1000.fa \
-s dme \
-t fasta:pro \
-o AROS1000
Command Explained

sudo docker run: tells docker to run

–rm: removes the container when the analysis has finished. The image will remain for future use.

-v /home/amcooksey/i5k/seq_pep:/seq_pep: tells docker to mount the ‘seq_pep’ directory I downloaded to the host machine to the ‘/seq_pep’ directory within the container. The syntax for this is: <absolute path on host>:<absolute path in container>

-v /home/amcooksey/i5k/sqlite3:/sqlite3: mounts ‘sqlite3’ directory on host machine into ‘sqlite3’ directory inside the container

-v $(pwd):/work-dir: mounts my current working directory on the host machine to ‘/work-dir’ in the container

agbase/kobas:3.0.3_0: the name of the Docker image to use

Tip

All the options supplied after the image name are KOBAS options

-a: Tells KOBAS to run the ‘annotate’ process.

-i AROS1000.fa: input file (peptide FASTA)

-s dme: Enter the species for the species of the sequences in your input file.

Note

If you don’t know the code for your species it can be found here: https://www.kegg.jp/kegg/catalog/org_list.html

If your species of interest is not available then you should choose the code for the closest-related species available

-t: input file type; in this case, protein FASTA.

-o AROS1000: name of output file

For information on output files see Understanding Your Results: Annotate

Identify Example Command

sudo docker run \
--rm \
-v /home/amcooksey/i5k/seq_pep:/seq_pep \
-v /home/amcooksey/i5k/sqlite3:/sqlite3 \
-v $(pwd):/work-dir \
agbase/kobas:3.0.3_0 \
-g \
-f AROS1000 \
-b dme \
-o ident_out
Command Explained

sudo docker run: tells docker to run

–rm: removes the container when the analysis has finished. The image will remain for future use.

-v /home/amcooksey/i5k/seq_pep:/seq_pep: tells docker to mount the ‘seq_pep’ directory I downloaded to the host machine to the ‘/seq_pep’ directory within the container. The syntax for this is: <absolute path on host>:<absolute path in container>

-v /home/amcooksey/i5k/sqlite3:/sqlite3: mounts ‘sqlite3’ directory on host machine into ‘go_info’ directory inside the container

-v $(pwd):/work-dir: mounts my current working directory on the host machine to ‘/work-dir’ in the container

agbase/kobas:3.0.3_0: the name of the Docker image to use

Tip

All the options supplied after the image name are KOBAS options

-g: Tells KOBAS to runt he ‘identify’ process.

-f AROS1000: output file from KOBAS annotate

-b dme: background; enter the species code for the species of the sequences in your input file.

Note

If you don’t know the code for your species it can be found here: https://www.kegg.jp/kegg/catalog/org_list.html

If your species of interest is not available then you should choose the code for the closest-related species available

-o ident_out: name of output file

For information on outputs see Understanding Your Results: Identify

Annotate and Identify Pipeline Example Command

sudo docker run \
--rm \
-v /home/amcooksey/i5k/seq_pep:/seq_pep \
-v /home/amcooksey/i5k/sqlite3:/sqlite3 \
-v $(pwd):/work-dir \
agbase/kobas:3.0.3_0 \
-j
-i AROS1000.fa \
-s dme \
-t fasta:pro
-o AROS1000
Command Explained

sudo docker run: tells docker to run

–rm: removes the container when the analysis has finished. The image will remain for future use.

-v /home/amcooksey/i5k/seq_pep:/seq_pep: tells docker to mount the ‘seq_pep’ directory I downloaded to the host machine to the ‘/seq_pep’ directory within the container. The syntax for this is: <absolute path on host>:<absolute path in container>

-v /home/amcooksey/i5k/sqlite3:/sqlite3: mounts ‘sqlite3’ directory on host machine into ‘go_info’ directory inside the container

-v $(pwd):/work-dir: mounts my current working directory on the host machine to ‘/work-dir’ in the container

agbase/kobas:3.0.3_0: the name of the Docker image to use

Tip

All the options supplied after the image name are KOBAS options

-j: Tells KOBAS to runt he ‘annotate’ process.

-i AROS1000.fa: input file (peptide FASTA)

-s dme: Enter the species for the species of the sequences in your input file.

Note

If you don’t know the code for your species it can be found here: https://www.kegg.jp/kegg/catalog/org_list.html

If your species of interest is not available then you should choose the code for the closest-related species available

-t: input file type; in this case, protein FASTA.

-o AROS1000: basename of output files

Note

This pipeline will automatically use the output of ‘annotate’ as the -f foreground input for ‘identify. This will also use your species option as the -b background input for ‘identify’.

For more information on outputs see Understanding Your Results: Annotate and Identify

Running KOBAS using Singularity

About Singularity

  • does not require ‘root’ permissions
  • runs all containers as the user that is logged into the host machine
  • HPC systems are likely to have Singularity installed and are unlikely to object if asked to install it (no guarantees).
  • can be run on any machine where is is installed
  • more information about installing Singularity
  • This tool was tested using Singularity 3.0. Users with Singularity 2.x will need to modify the commands accordingly.

HPC Job Schedulers

Although Singularity can be installed on any computer this documentation assumes it will be run on an HPC system. The tool was tested on a PBSPro system and the job submission scripts below reflect that. Submission scripts will need to be modified for use with other job scheduler systems.

Getting the KOBAS container

The KOBAS tool is available as a Docker container on Docker Hub: KOBAS container

The container can be pulled with this command:

singularity pull docker://agbase/kobas:3.0.3_0

Running KOBAS with Data

Getting the Help and Usage Statement

Example PBS script:

#!/bin/bash
#PBS -N kobas
#PBS -W group_list=fionamcc
#PBS -l select=1:ncpus=28:mem=168gb
#PBS -q standard
#PBS -l walltime=6:0:0
#PBS -l cput=168:0:0

module load singularity

cd /rsgrps/shaneburgess/amanda/i5k/kobas

singularity pull docker://agbase/kobas:3.0.3_0

singularity run \
kobas_3.0.3_0.sif \
-h

See Help and Usage Statement

Tip

There are 3 directories built into this container. These directories should be used to mount data.

  • /seq_pep
  • /sqlite3
  • /work-dir

Example PBS Script for Annotate Process

#!/bin/bash
#PBS -N kobas
#PBS -W group_list=fionamcc
#PBS -l select=1:ncpus=28:mem=168gb
#PBS -q standard
#PBS -l walltime=6:0:0
#PBS -l cput=168:0:0

module load singularity

cd /rsgrps/shaneburgess/amanda/i5k/kobas

singularity pull docker://agbase/kobas:3.0.3_0

singularity run \
-B /rsgrps/shaneburgess/amanda/i5k/seq_pep:/seq_pep \
-B /rsgrps/shaneburgess/amanda/i5k/sqlite3:/sqlite3 \
-B /rsgrps/shaneburgess/amanda/i5k/kobas:/work-dir \
kobas_3.0.3_0.sif \
-a
-i AROS1000.fa \
-s dme \
-t fasta:pro \
-o AROS1000 \
Command Explained

singularity run: tells Singularity to run

-B /rsgrps/shaneburgess/amanda/i5k/seq_pep:/seq_pep: tells docker to mount the ‘seq_pep’ directory I downloaded to the host machine to the ‘/seq_pep’ directory within the container. The syntax for this is: <absolute path on host>:<absolute path in container>

-B /rsgrps/shaneburgess/amanda/i5k/sqlite3:/sqlite3: mounts ‘sqlite3’ directory on host machine into ‘go_info’ directory inside the container

-B /rsgrps/shaneburgess/amanda/i5k/kobas:/work-dir: mounts my current working directory on the host machine to ‘/work-dir’ in the container

kobas_3.0.3_0.sif: the name of the Singularity image to use

Tip

All the options supplied after the image name are KOBAS options

-a: Tells KOBAS to runt he ‘annotate’ process.

-i AROS1000.fa: input file (peptide FASTA)

-s dme: Enter the species for the species of the sequences in your input file.

Note

If you don’t know the code for your species it can be found here: https://www.kegg.jp/kegg/catalog/org_list.html

If your species of interest is not available then you should choose the code for the closest-related species available

-t: input file type; in this case, protein FASTA.

-o AROS1000: name of output file

For information on output files see Understanding Your Results: Annotate

Example PBS Script for Identify Process

#!/bin/bash
#PBS -N kobas
#PBS -W group_list=fionamcc
#PBS -l select=1:ncpus=28:mem=168gb
#PBS -q standard
#PBS -l walltime=6:0:0
#PBS -l cput=168:0:0

module load singularity

cd /rsgrps/shaneburgess/amanda/i5k/kobas

singularity pull docker://agbase/kobas:3.0.3_0

singularity run \
-B /rsgrps/shaneburgess/amanda/i5k/seq_pep:/seq_pep \
-B /rsgrps/shaneburgess/amanda/i5k/sqlite3:/sqlite3 \
-B /rsgrps/shaneburgess/amanda/i5k/kobas:/work-dir \
kobas_3.0.3_0.sif \
-g
-f AROS1000 \
-b dme \
-o ident_out
Command Explained

singularity run: tells Singularity to run

-B /rsgrps/shaneburgess/amanda/i5k/seq_pep:/seq_pep: tells docker to mount the ‘seq_pep’ directory I downloaded to the host machine to the ‘/seq_pep’ directory within the container. The syntax for this is: <absolute path on host>:<absolute path in container>

-B /rsgrps/shaneburgess/amanda/i5k/sqlite3:/sqlite3: mounts ‘sqlite3’ directory on host machine into ‘go_info’ directory inside the container

-B /rsgrps/shaneburgess/amanda/i5k/kobas:/work-dir: mounts my current working directory on the host machine to ‘/work-dir’ in the container

kobas_3.0.3_0.sif: the name of the Singularity image to use

Tip

All the options supplied after the image name are KOBAS options

-g: Tells KOBAS to runt he ‘identify’ process.

-f AROS1000: output file from ‘annotate’

-b dme: background; enter the species for the species of the sequences in your input file.

Note

If you don’t know the code for your species it can be found here: https://www.kegg.jp/kegg/catalog/org_list.html

If your species of interest is not available then you should choose the code for the closest-related species available

-o ident_out: name of output file

For information on output see Understanding Your Results: Identify

Example PBS Script for Annotate and Identify Pipeline

#!/bin/bash
#PBS -N kobas
#PBS -W group_list=fionamcc
#PBS -l select=1:ncpus=28:mem=168gb
#PBS -q standard
#PBS -l walltime=6:0:0
#PBS -l cput=168:0:0

module load singularity

cd /rsgrps/shaneburgess/amanda/i5k/kobas

singularity pull docker://agbase/kobas:3.0.3_0

singularity run \
-B /rsgrps/shaneburgess/amanda/i5k/seq_pep:/seq_pep \
-B /rsgrps/shaneburgess/amanda/i5k/sqlite3:/sqlite3 \
-B /rsgrps/shaneburgess/amanda/i5k/kobas:/work-dir \
kobas_3.0.3_0.sif \
-j
-i AROS1000.fa \
-s dme \
-t fasta:pro \
-o AROS1000 \
Command Explained

singularity run: tells Singularity to run

-B /rsgrps/shaneburgess/amanda/i5k/seq_pep:/seq_pep: tells docker to mount the ‘seq_pep’ directory I downloaded to the host machine to the ‘/seq_pep’ directory within the container. The syntax for this is: <absolute path on host>:<absolute path in container>

-B /rsgrps/shaneburgess/amanda/i5k/sqlite3:/sqlite3: mounts ‘sqlite3’ directory on host machine into ‘go_info’ directory inside the container

-B /rsgrps/shaneburgess/amanda/i5k/kobas:/work-dir: mounts my current working directory on the host machine to ‘/work-dir’ in the container

kobas_3.0.3_0.sif: the name of the Singularity image to use

Tip

All the options supplied after the image name are KOBAS options

-j: Tells KOBAS to runt he ‘annotate’ process.

-i AROS1000.fa: input file (peptide FASTA)

-s dme: Enter the species for the species of the sequences in your input file.

Note

If you don’t know the code for your species it can be found here: https://www.kegg.jp/kegg/catalog/org_list.html

If your species of interest is not available then you should choose the code for the closest-related species available

-t: input file type; in this case, protein FASTA.

-o AROS1000: name of output file

Note

This pipeline will automatically use the output of ‘annotate’ as the -f foreground input for ‘identify’. This will also use your species option as the -b background input for ‘identify’.

For information on outputs see Understanding Your Results: Annotate and Identify

Understanding Your Results

Annotate

If all goes well, you should get the following:

  • seq_pep folder: This folder contains the BLAST database files used in your analysis.
  • sqlite3 folder: This folder contains the annotation database files used in your analysis
  • <species>.tsv: This is the tab-delimited output from the BLAST search. It is unlikely that you will need to look at this file.
  • <output_file_name_you_provided>: KOBAS-annotate generates a text file with the name you provide. It has two sections.

The first sections looks like this:

##dme       Drosophila melanogaster (fruit fly)
##Method: BLAST     Options: evalue <= 1e-05
##Summary:  87 succeed, 0 fail

#Query      Gene ID|Gene name|Hyperlink
lcl|NW_020311285.1_prot_XP_012256083.1_15   dme:Dmel_CG34349|Unc-13-4B|http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG34349
lcl|NW_020311286.1_prot_XP_020708336.1_46   dme:Dmel_CG6963|gish|http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG6963
lcl|NW_020311285.1_prot_XP_020707987.1_39   dme:Dmel_CG30403||http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG30403

The second section follows a dashed line and looks like this:

--------------------

////
Query:                      lcl|NW_020311285.1_prot_XP_012256083.1_15
Gene:                       dme:Dmel_CG34349        Unc-13-4B
Entrez Gene ID:             43002
////
Query:                      lcl|NW_020311286.1_prot_XP_020708336.1_46
Gene:                       dme:Dmel_CG6963 gish
Entrez Gene ID:             49701
Pathway:                    Hedgehog signaling pathway - fly        KEGG PATHWAY    dme04341
////
Query:                      lcl|NW_020311285.1_prot_XP_020707987.1_39
Gene:                       dme:Dmel_CG30403
Entrez Gene ID:             246595
////
Query:                      lcl|NW_020311285.1_prot_XP_020707989.1_40
Gene:                       dme:Dmel_CG6148 Past1
Entrez Gene ID:             41569
Pathway:                    Endocytosis     KEGG PATHWAY    dme04144
                            Hemostasis      Reactome        R-DME-109582
                            Factors involved in megakaryocyte development and platelet production   Reactome        R-DME-98323

Identify

If all goes well, you should get the following:

  • sqlite3 folder: This folder contains the annotation database files used in your analysis
  • <output_file_name_you_provided>: KOBAS identify generates a text file with the name you provide.
##Databases: PANTHER, KEGG PATHWAY, Reactome, BioCyc
##Statistical test method: hypergeometric test / Fisher's exact test
##FDR correction method: Benjamini and Hochberg

#Term       Database        ID      Input number    Background number       P-Value Corrected P-Value       Input   Hyperlink
Hedgehog signaling pathway - fly    KEGG PATHWAY    dme04341        12      33      3.20002656734e-18       1.76001461204e-16       lcl|NW_020311286.1_prot_XP_012256678.1_51|lcl|NW_020311286.1_prot_XP_025602973.1_48|lcl|NW_020311286.1_prot_XP_012256683.1_52|lcl|NW_020311286.1_prot_XP_012256679.1_55|lcl|NW_020311286.1_prot_XP_012256674.1_54|lcl|NW_020311286.1_prot_XP_020708336.1_46|lcl|NW_020311285.1_prot_XP_012256108.1_32|lcl|NW_020311286.1_prot_XP_012256682.1_53|lcl|NW_020311286.1_prot_XP_025603025.1_47|lcl|NW_020311286.1_prot_XP_020708334.1_49|lcl|NW_020311285.1_prot_XP_012256109.1_33|lcl|NW_020311286.1_prot_XP_020708333.1_50 http://www.genome.jp/kegg-bin/show_pathway?dme04341/dme:Dmel_CG6963%09red/dme:Dmel_CG6054%09red
Hedgehog signaling pathway  PANTHER P00025  6       13      3.6166668094e-10        9.94583372585e-09       lcl|NW_020311286.1_prot_XP_025602279.1_78|lcl|NW_020311286.1_prot_XP_025602289.1_76|lcl|NW_020311286.1_prot_XP_025602264.1_79|lcl|NW_020311285.1_prot_XP_012256108.1_32|lcl|NW_020311285.1_prot_XP_012256109.1_33|lcl|NW_020311286.1_prot_XP_012256943.1_77     http://www.pantherdb.org/pathway/pathwayDiagram.jsp?catAccession=P00025
Signaling by NOTCH2 Reactome        R-DME-1980145   3       8       2.00259649553e-05       0.000275357018136       lcl|NW_020311285.1_prot_XP_012256118.1_28|lcl|NW_020311285.1_prot_XP_012256117.1_27|lcl|NW_020311285.1_prot_XP_012256119.1_26   http://www.reactome.org/cgi-bin/eventbrowser_st_id?ST_ID=R-DME-1980145

Annotate and Identify Pipeline

If all goes well, you should get the following:

  • logs folder: This folder contains the ‘conder_stderr’ and ‘condor_stdout’ files. The files record feedback, progress and, importantly, any errors the app encountered during the analysis. You won’t normally need to look at these but they are very helpful in figuring out what may have happened if your output doesn’t look like you expected.
  • sqlite3 folder: This folder contains the annotation database files used in your analysis
  • seq_pep folder: This folder contains the BLAST database files used in your analysis.
  • <species>.tsv: This is the tab-delimited output from the BLAST search. It is unlikely that you will need to look at this file.
  • <basename>_annotate_out.txt: KOBAS annotate generates a text file with the name you provide. It has two sections.

The first sections looks like this:

##dme       Drosophila melanogaster (fruit fly)
##Method: BLAST     Options: evalue <= 1e-05
##Summary:  87 succeed, 0 fail

#Query      Gene ID|Gene name|Hyperlink
lcl|NW_020311285.1_prot_XP_012256083.1_15   dme:Dmel_CG34349|Unc-13-4B|http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG34349
lcl|NW_020311286.1_prot_XP_020708336.1_46   dme:Dmel_CG6963|gish|http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG6963
lcl|NW_020311285.1_prot_XP_020707987.1_39   dme:Dmel_CG30403||http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG30403

The second section follows a dashed line and looks like this:

--------------------

////
Query:                      lcl|NW_020311285.1_prot_XP_012256083.1_15
Gene:                       dme:Dmel_CG34349        Unc-13-4B
Entrez Gene ID:             43002
////
Query:                      lcl|NW_020311286.1_prot_XP_020708336.1_46
Gene:                       dme:Dmel_CG6963 gish
Entrez Gene ID:             49701
Pathway:                    Hedgehog signaling pathway - fly        KEGG PATHWAY    dme04341
////
Query:                      lcl|NW_020311285.1_prot_XP_020707987.1_39
Gene:                       dme:Dmel_CG30403
Entrez Gene ID:             246595
////
Query:                      lcl|NW_020311285.1_prot_XP_020707989.1_40
Gene:                       dme:Dmel_CG6148 Past1
Entrez Gene ID:             41569
Pathway:                    Endocytosis     KEGG PATHWAY    dme04144
                            Hemostasis      Reactome        R-DME-109582
                            Factors involved in megakaryocyte development and platelet production   Reactome        R-DME-98323mel_CG6963
lcl|NW_020311285.1_prot_XP_020707987.1_39   dme:Dmel_CG30403||http://www.genome.jp/dbget-bin/www_bget?dme:Dmel_CG30403
  • <basename>_identify_out.txt: KOBAS identify generates a text file with the name you provide.
##Databases: PANTHER, KEGG PATHWAY, Reactome, BioCyc
##Statistical test method: hypergeometric test / Fisher's exact test
##FDR correction method: Benjamini and Hochberg

#Term       Database        ID      Input number    Background number       P-Value Corrected P-Value       Input   Hyperlink
Hedgehog signaling pathway - fly    KEGG PATHWAY    dme04341        12      33      3.20002656734e-18       1.76001461204e-16       lcl|NW_020311286.1_prot_XP_012256678.1_51|lcl|NW_020311286.1_prot_XP_025602973.1_48|lcl|NW_020311286.1_prot_XP_012256683.1_52|lcl|NW_020311286.1_prot_XP_012256679.1_55|lcl|NW_020311286.1_prot_XP_012256674.1_54|lcl|NW_020311286.1_prot_XP_020708336.1_46|lcl|NW_020311285.1_prot_XP_012256108.1_32|lcl|NW_020311286.1_prot_XP_012256682.1_53|lcl|NW_020311286.1_prot_XP_025603025.1_47|lcl|NW_020311286.1_prot_XP_020708334.1_49|lcl|NW_020311285.1_prot_XP_012256109.1_33|lcl|NW_020311286.1_prot_XP_020708333.1_50 http://www.genome.jp/kegg-bin/show_pathway?dme04341/dme:Dmel_CG6963%09red/dme:Dmel_CG6054%09red
Hedgehog signaling pathway  PANTHER P00025  6       13      3.6166668094e-10        9.94583372585e-09       lcl|NW_020311286.1_prot_XP_025602279.1_78|lcl|NW_020311286.1_prot_XP_025602289.1_76|lcl|NW_020311286.1_prot_XP_025602264.1_79|lcl|NW_020311285.1_prot_XP_012256108.1_32|lcl|NW_020311285.1_prot_XP_012256109.1_33|lcl|NW_020311286.1_prot_XP_012256943.1_77     http://www.pantherdb.org/pathway/pathwayDiagram.jsp?catAccession=P00025
Signaling by NOTCH2 Reactome        R-DME-1980145   3       8       2.00259649553e-05       0.000275357018136       lcl|NW_020311285.1_prot_XP_012256118.1_28|lcl|NW_020311285.1_prot_XP_012256117.1_27|lcl|NW_020311285.1_prot_XP_012256119.1_26   http://www.reactome.org/cgi-bin/eventbrowser_st_id?ST_ID=R-DME-1980145

Contact us.