Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/warelab/gramene-ensembl

Gramene's Ensembl plugins, extensions, configuration.
https://github.com/warelab/gramene-ensembl

gramene perl

Last synced: about 1 month ago
JSON representation

Gramene's Ensembl plugins, extensions, configuration.

Awesome Lists containing this project

README 

        
# GRAMENE BLAST; How to set up the blast server

#==============================================



# These instructions were tested for the versions below on 

# catskill.cshl.edu 



# CODE INSTALLATION

#------------------



# bash shell assumed

# Set code/data versions (will change with each release)

# These are just used for convenience for this README

export ENSVER=65

export GRAVER=34b

export EGVER=12



# Fetch the version-specific code into a clean directory

mkdir /usr/local/svntmp/ensembl-grmblast-$ENSVER

cd /usr/local/svntmp/ensembl-grmblast-$ENSVER



cvs -d :pserver:[email protected]:/cvsroot/ensembl login

[pass CVSUSER]



cvs -d :pserver:[email protected]:/cvsroot/ensembl \

co -r branch-ensembl-$ENSVER ensembl-api



cvs -d :pserver:[email protected]:/cvsroot/ensembl \

co -r branch-ensembl-$ENSVER ensembl-website



cvs -d :pserver:[email protected]:/cvsroot/ensembl \

co -r branch-ensemblgenomes-$EGVER-$ENSVER eg-plugins



svn co http://svn.warelab.org/gramene/branches/build${GRAVER}-branch \

gramene-$GRAVER



# Move the code directory into place

sudo mv /usr/local/svntmp/ensembl-grmblast-$ENSVER /usr/local



# Create links to code that does not change

cd /usr/local/ensembl-grmblast-$ENSVER 

ln -s gramene-$GRAVER gramene-live

ln -s /usr/local/bioperl-1-2-3 bioperl-live

ln -s /usr/local/biomart-perl-0_6 biomart-perl

ln -s /usr/local/apache2 apache2

cp /usr/local/vcftools_0.1.5/lib/Vcf.pm /usr/local/ensembl-grmblast/modules/



# Create temporary dirs

mkdir logs

mkdir -p /usr/local/ensembl-grmblast/htdocs/BlastView/minified



# Move the Ensembl Plugins file into place

cp gramene-live/ensembl-plugins/grmblast/conf/Plugins.pm conf/



# Stop the old server, and start the new

# NOTE: the new server should have been tested first!



sudo kill `cat /usr/local/ensembl-grmblast/logs/httpd.pid`

sudo rm /usr/local/ensembl-grmblast

sudo ln -s /usr/local/ensembl-grmblast-$ENSVER /usr/local/ensembl-grmblast

sudo /etc/init.d/apachectl-grmblast start



# DATA INSTALLATION

#------------------



# Get the version-specific FASTA databases

# Note that we download these from the EnsemblGenomes FTP site, 

# But they can be generated direct from the Ensembl databases

# using the utils/do_fasta_dump script



##get sequences

# for EG genomes, we can download from EG ftp site



wget -nH -r --cut-dirs=3 \

"ftp://ftp.ensemblgenomes.org/pub/release-$EGVER/plants/fasta/"



# for Gramene specific genomes such as OGE genomes, create locally

cd ensembl-live/

perl  utils/do_fasta_dump  

	--no_ssaha --species $s 

	-type dna_toplevel -type cdna_all -type pep_all -type ncrna 

	-dumpdir OGEdump/ 

	-release 71 

	--no_log 

	-email [email protected];



# Index the FASTA DBs

find fasta/ -name \*.gz -exec gunzip {} \;

find fasta/ -name \*.pep.\*.fa -exec /usr/local/wublast/xdformat -p {} \;

find fasta/ -name \*dna\*.fa -exec /usr/local/wublast/xdformat -n {} \;



# Move the indexes into place, and update the symlink

mkdir /usr/local/blastdb/gramene-$ENSVER

find fasta/ -name \*.fa.x?? -exec mv {} /usr/local/blastdb/gramene-$ENSVER \;

sudo rm /usr/local/blastdb/gramene

sudo ln -s /usr/local/blastdb/gramene-$ENSVER \

/usr/local/blastdb/gramene



# Some slight renaming to convert from EG to Ensembl convention

cd /usr/local/blastdb/gramene

ls | perl -e 'while(<>){ chomp; my $ori=$_; my $new=$ori; if( $ori =~ /^.*\..*\.(.*\.).*\..*\.fa/..*$ ){ $new =~ s/$1//; `mv $ori $new` } }'

For example, the filename should be Oryza_sativa.IRGSP-1.0.dna.toplevel.fa.* or Zea_mays.AGPv3.pep.all.fa.*

The format should be, and we only need *.dna.toplevel.fa.* and *.pep.all.fa.*, we do not need the individual chromosome files to be indexed,

Species.assemblyVersion.dna.toplevel.fa.*

Species.assemblyVersion.pep.all.fa.*



# MAINTENANCE

#------------



Ensure that there is a startup script in /etc/init.d that is correctly

linked from /etc/rc.d to ensure that the web server is restarted on

server reboot.



There are also some scripts that need installing in the root user's

crontab that clean up temporary blast files and database records. Here

is an example (please replace  with the value of the $ENSROOT

variable);



---

# Clean ensembl* tmp and img-tmp files, and the blast database

0  4 * * * find /*tmp -type f -atime +1 -exec rm {} \; > /dev/null 2>&1

0  4 * * * find /img-cache -type f -atime +1 -exec rm {} \; > /dev/null 2>&1

0  3 * * * perl /utils/blast_cleaner.pl 

---