Install coge

Installing CoGe on Ubuntu

Note: these instructions were last updated and verified on June 3rd, 2016.

Initial Dependencies

Run the following command:

sudo apt-get -y install {package}

where {package} is each of the following:

apache2
aragorn
blast2
build-essential
checkinstall
expat
gcc-multilib
git
graphviz
imagemagick
libdb-dev
libgd2-xpm-dev
libperl-dev
libgd-gd2-perl
libconfig-yaml-perl
libssl-dev
libzmq3-dev
mysql-server
ncbi-blast+
ncbi-blast+-legacy
njplot
phpmyadmin
python-dev
python-numpy
python-software-properties
samtools
swig
sqlite3
ttf-mscorefonts-installer
ubuntu-dev-tools
libapache-asp-perl
libapache2-mod-perl2
libapache2-mod-wsgi
python-pip
nodejs
npm
libboost-all-dev  (for TopHat)

Create MySQL database

Dump CoGe database schema (if using existing CoGe installation, otherwise see schema file below).

mysqldump -d -h localhost -u root -pXXXXXXX coge | sed 's/AUTO_INCREMENT=[0-9]*\b//' > coge_mysql_schema.sql

• CoGe MySQL database schema file (updated June 3rd, 2016): http://genomevolution.org/coge/coge_mysql_schema.sql
• Note: be sure to disable AppArmor for MySQL.

Create new CoGe Database

create database coge

Initialize new coge database

mysql -u root -pXXXXXXXX coge < coge_mysql_schema.sql

Populate a few entries in the feature_type table

• Use the table here which contains 10 feature types: http://genomevolution.org/coge/coge_feature_types.sql

mysql -u root -pXXXXXXXXX coge < coge_feature_types.sql

Create new MySQL user for the CoGe database

use mysql;
create user 'coge'@'localhost' IDENTIFIED BY 'XXXXXX';
grant all privileges on coge.* to coge;
flush privileges;

Note: The CoGe web-user needs edit/insert permission on some tables. Here is a snapshot of what these are:

Deploy the Web Site

Generate a public key and add to your GitHub account

See https://help.github.com/articles/generating-an-ssh-key/

Download the CoGe repository

git clone https://github.com/LyonsLab/coge.git

Run setup script to make required subdirectories

cd coge/web
./setup.sh

Configure apache

The /etc/apache2/sites-available/default.conf should look like this:

<VirtualHost *>
	ServerAdmin webmasterl@localhost
	DocumentRoot /opt/coge/web

	<Files *.pl>
    		SetHandler perl-script
    		PerlResponseHandler ModPerl::Registry
    		Options +ExecCGI
    		PerlSendHeader On
	</Files>

	<Directory />
		Options FollowSymLinks
		AllowOverride None
	</Directory>

	Alias /gobe/ /opt/coge/web/gobe/
	<Directory /opt/coge/web/gobe/>
		Options +FollowSymLinks +ExecCGI
		AddHandler wsgi-script .py
	</Directory>

	<Directory /opt/coge>
		Options Includes ExecCGI FollowSymLinks
    		AllowOverride All
    		SetEnv COGE_HOME "/opt/coge/"
    		Order allow,deny
    		Allow from all
	</Directory>

	<Directory /opt/coge/web/services/>
        	Options +FollowSymLinks +ExecCGI
        	AddHandler wsgi-script .py
 	</Directory>

 	<Directory /opt/coge/web/services/JBrowse/JBrowse_TrackContent_WS/>
        	Options +FollowSymLinks +ExecCGI
        	AddHandler wsgi-script .py
  	</Directory>
	ScriptAliasMatch (?i)^/coge/jex(.*) /opt/coge/web/services/jex.py/$1
        AliasMatch (?i)^/coge(.*) /opt/coge/web/$1
        ProxyPass /coge/api/v1/ http://localhost:3303/
        ProxyPassReverse /coge/api/v1/ http://localhost:3303/

	ErrorLog /var/log/apache2/error.log

	# Possible values include: debug, info, notice, warn, error, crit, alert, emerg.
	LogLevel warn

	CustomLog /var/log/apache2/access.log combined
	ServerSignature On
</VirtualHost>

Enable Required Apache Modules

sudo a2enmod rewrite headers proxy proxy_http expires perl ssl

and reset Apache

Configure coge.conf file

Replacing XXX's with your own information. (Change paths as necessary; this template is configured for having the Coge directory in the path: /opt/coge)

#
# CoGe Configuration File
#

#database configuration
DB      mysql
DBNAME	coge
DBHOST	localhost
DBPORT	3307
DBUSER	coge
DBPASS	321coge123

#basic auth name and password
AUTHNAME 6mv9x9lyts8oje8uj3t6yo
AUTHPASS f59ba33ee35d363ffefd8b27b375e587b0e5c7a1

# DE public key for JWT in resources/
DE_PUBLIC_KEY DE_rsa.pub

#web cookie name
COOKIE_NAME cogec

#support email address
SUPPORT_EMAIL coge.genome@gmail.com

#data dir for coge's programs
DATADIR	/storage/coge/data/

#cache dir
CACHEDIR /scratch/coge/cache

#dir for pair-wise whole genome comparisons (e.g. SynMap)
DIAGSDIR /opt/apache2/coge/web/data/diags/

#dir for popgen analysis results
POPGENDIR /storage/coge/data/popgen/

#fasta dir
FASTADIR /opt/apache2/coge/web/data/fasta/

#sequence dir
SEQDIR /storage/coge/data/genomic_sequence/

#experiment dir
EXPDIR /storage/coge/data/experiments/

#temp dir for coge
TEMPDIR /opt/apache2/coge/web/tmp/

#secure temp dir
SECTEMPDIR /scratch/coge/tmp/

# IRODS dir
IRODSDIR /iplant/home/<USER>/coge_data
IRODSSHARED /iplant/home/shared
IRODSENV /opt/apache/coge/irodsEnv

#Base URL for web-site
URL /coge/
API_URL /api/v1/

#URL for temp directory
TEMPURL /coge/tmp/

#blast style scoring matrix dirs
BLASTMATRIX /storage/coge/data/blast/matrix/

#blastable DB
BLASTDB /scratch/coge/cache/blast/db/

#lastable DB
LASTDB /scratch/coge/cache/last/db/

#directory for bed files
BEDDIR /opt/apache2/coge/web/data/bed/

#WIKI URL
WIKI_URL https://genomevolution.org/wiki/index.php

#servername for links
#SERVER https://genomevolution.org/coge/
SERVER http://10.140.65.127/coge/

#CAS URL
CAS_URL https://auth.iplantcollaborative.org/cas4
USER_API_URL https://agave.iplantc.org:443/profiles/v2
MOJOLICIOUS_PORT 3303

# Job Engine Server
JOBSERVER localhost

# Job Engine Port
JOBPORT 5151

#directory for caching genome browser images
IMAGE_CACHE /opt/apache2/coge/web/data/image_cache/

#maximum number of processor to use for multi-CPU systems
MAX_PROC 44
COGE_BLAST_MAX_PROC 8

#True Type Font
FONT /usr/local/fonts/arial.ttf

#various programs
BL2SEQ                 /usr/local/bin/legacy_blast.pl bl2seq
BLAST                  /usr/local/bin/legacy_blast.pl blastall
MULTI_LASTZ            /opt/apache2/coge/bin/blastz_wrapper/blastz.py
LAST_PATH              /opt/apache2/coge/bin/last_wrapper/
MULTI_LAST             /opt/apache2/coge/bin/last_wrapper/last.py
LAGAN                  /opt/apache2/coge/bin/lagan-64bit/lagan.pl
LAGANDIR               /opt/apache2/coge/bin/lagan-64bit/
CHAOS                  /opt/apache2/coge/bin/lagan-64bit/chaos
GENOMETHREADER         /opt/apache2/coge/bin/gth
DIALIGN                /opt/apache2/coge/bin/dialign2_dir/dialign2-2_coge
DIALIGN2	       /opt/apache2/coge/bin/dialign2_dir/dialign2-2_coge
DIALIGN2_DIR	       /opt/apache2/coge/bin/dialign2_dir/
HISTOGRAM	       /opt/apache2/coge/bin/histogram.pl
KS_HISTOGRAM	       /opt/apache2/coge/bin/ks_histogram.pl
TANDEM_FINDER	       /opt/apache2/coge/bin/dagchainer/tandems.py
DAGCHAINER	       /opt/apache2/coge/bin/dagchainer_bp/dag_chainer.py
EVALUE_ADJUST	       /opt/apache2/coge/bin/dagchainer_bp/dagtools/evalue_adjust.py
FIND_NEARBY	       /opt/apache2/coge/bin/dagchainer_bp/dagtools/find_nearby.py
QUOTA_ALIGN	       /opt/apache2/coge/bin/quota-alignment/quota_align.py
CLUSTER_UTILS	       /opt/apache2/coge/bin/quota-alignment/cluster_utils.py
BLAST2RAW	       /opt/apache2/coge/bin/quota-alignment/scripts/blast_to_raw.py
SYNTENY_SCORE	       /opt/apache2/coge/bin/quota-alignment/scripts/synteny_score.py
CODEML	               /opt/apache2/coge/bin/codeml/codeml-coge
CODEMLCTL              /opt/apache2/coge/bin/codeml/codeml.ctl
CONVERT_BLAST          /opt/apache2/coge/bin/convert_long_blast_to_short_blast_names.pl
DATASETGROUP2BED       /opt/apache2/coge/bin/dataset_group_2_bed.pl

#stuff for Mauve and whole genome alignments
MAUVE        /opt/apache2/coge/bin/GenomeAlign/progressiveMauve-muscleMatrix
COGE_MAUVE   /opt/apache2/coge/bin/GenomeAlign/mauve_alignment.pl
MAUVE_MATRIX /opt/apache2/coge/web/data/blast/matrix/nt/Mauve-Matrix-GenomeAlign

# RNA-seq pipelines
PARSE_CUFFLINKS /opt/apache2/coge/scripts/parse_cufflinks.py

# SNP pipelines
PLATYPUS /opt/apache2/coge/bin/Platypus_0.8.1/Platypus.py
GATK     /opt/apache2/coge/bin/GenomeAnalysisTK.jar
PICARD   /opt/apache2/coge/bin/picard-tools-2.4.1/picard.jar

# ChIP-seq pipeline
HOMER_DIR /opt/apache2/coge/bin/Homer

#THIRD PARTY URLS
GENFAMURL http://dev.gohelle.cirad.fr/genfam/?q=content/upload

Install Perl Modules

• Install cpanminus

sudo cpan install App::cpanminus

• Install third-party modules required by CoGe

cat modules.txt | xargs sudo cpanm

• Install CoGe-specific modules

./make_perl.sh

• After installing modules, reset the Apache webserver

sudo service apache2 restart

Install Python Modules

sudo pip install pyzmq

Install Javascript dependencies

• Install javascript dependencies

sudo ln -s /usr/bin/nodejs /usr/bin/node
sudo npm install -g bower
bower install

Install Third-Party Bioinformatics Tools

Download the programs listed below and follow the installation instructions on their respective websites.

Most programs can be installed with the following commands (but check the documentation for each program):

./configure --prefix=/usr/local/
make
sudo make install

• GSNAP/GMAP: http://research-pub.gene.com/gmap/
• FastBit: https://sdm.lbl.gov/fastbit/
• Clustalw: http://www.clustal.org/clustal2/
• Bowtie: http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
• TopHat: http://ccb.jhu.edu/software/tophat/index.shtml
• HISAT2: https://ccb.jhu.edu/software/hisat2/index.shtml
• Cufflinks: http://cole-trapnell-lab.github.io/cufflinks/
• Nwalign: https://pypi.python.org/pypi/nwalign/?
• Cutadapt: http://cutadapt.readthedocs.io/en/stable/installation.html ... sudo su ; sudo pip install cutadapt
• TrimGalore: http://www.bioinformatics.babraham.ac.uk/projects/trim_galore/
• Picard: http://broadinstitute.github.io/picard/ ... requires Java 8
• Platypus: http://www.well.ox.ac.uk/platypus
• HSTlib (required by Platypus): http://www.htslib.org/download/
• Lastz: download the tarball http://www.bx.psu.edu/~rsharris/lastz/ then edit the src/Makefile and remove the word -Werror from line 31. Then run make and make install.
• Last aligner (v731 or greater is required): http://last.cbrc.jp/
• VCFTools: https://github.com/vcftools/vcftools
• Vcfutils: https://github.com/lh3/samtools/blob/master/bcftools/vcfutils.pl
• EMBOSS (sizeseq program): http://emboss.sourceforge.net/ ... Run "sudo ldconfig" after "make install"
• iCommands: https://github.com/irods/irods-legacy (OPTIONAL: only required if CyVerse authentication and Data Store services are available; note that the legacy version is required, not the latest)
• Bismark: http://www.bioinformatics.babraham.ac.uk/projects/bismark/
• BWAmeth: https://github.com/brentp/bwa-meth
• Homer: http://homer.salk.edu/homer/ ... sudo perl ./configureHomer.pl -install homer
• Blat (required by Homer): https://genome.ucsc.edu/FAQ/FAQblat.html

Install Third-Party Fonts

Download from here: https://www.microsoft.com/typography/fonts/font.aspx?FMID=1705

And copy to /usr/local/fonts/arial.ttf (or whatever path you set in the coge.conf config file under FONT)

Install blast matrices

cd /storage/coge/data/blast
git clone https://github.com/LyonsLab/blast-matrix.git
mv blast-matrix matrix

Install JBrowse

Copy from existing CoGe installation if one exists. Otherwise, download and install the JBrowse package from http://jbrowse.org/install/

unzip JBrowse-1.11.4-dev.zip
mv JBrowse-1.11.4 /coge/web/js/jbrowse

Install CCTools

• Download a stable release that is 4.3 or greater from http://ccl.cse.nd.edu/software/downloadfiles.php

• Extract the file (this example is using version 4.3 which may differ from the version downloaded)

tar xzvf cctools-4.3.0-source.tar.gz

• Compile and install

cd cctools-4.3.0-source
./configure --prefix /usr/local
make
sudo make install

• Add the following upstart scripts for the work_queue_pool and catalog_server to /etc/init

By default the pool directory for work_queue will be in /storage/work_queue adjust the directory as needed.

# /etc/init/.conf

description "The cctools work queue pool"
author "Evan Briones"

start on (local-filesystems and net-device-up IFACE=eth0)

stop on shutdown

respawn limit 30 60

pre-start script
    POOL_DIR=/storage/work_queue
    LOG_FILE=$POOL_DIR/logs/work_queue_pool.log

    # Add the pool directory and set ownership
    if ! [ -d "$WORK_DIR" ]; then
        mkdir -p $POOL_DIR/workers
        mkdir -p $POOL_DIR/logs
        chown -R www-data:www-data $POOL_DIR
    fi

    # Remove the pidfile if it exists
    rm -f $POOL_DIR/work_queue_pool.pid

    # Archive old log and timestamp the value
    if [ -f "$LOG_FILE" ]; then
        TIMESTAMP=$(date  +"%Y-%m-%d.%H.%m.%S")
        mv -f $LOG_FILE "$LOGFILE.$TIMESTAMP"
    fi
end script

script
    POOL_DIR=/storage/work_queue
    LOG_FILE=$POOL_DIR/logs/work_queue_pool.log
    #CONFIG=/etc/yerba/work_queue_pool.conf
    #WORK_QUEUE_POOL=$(which work_queue_pool)
    WORK_QUEUE_FACTORY=$(which work_queue_factory)

    export CATALOG_HOST=localhost
    export CATALOG_PORT=1024

    exec start-stop-daemon -c www-data -g www-data -d $POOL_DIR --start \
        -p $POOL_DIR/work_queue_pool.pid --exec $WORK_QUEUE_FACTORY \
        -- -T local -M coge-main -d all -o $LOG_FILE -w 10 \
        -S $POOL_DIR -E "--workdir=$POOL_DIR/workers"
end script

# /etc/init/.conf

description "The cctools catalog server"
author "Evan Briones"

start on (local-filesystems and net-device-up IFACE=eth0)

stop on shutdown

respawn limit 30 60

script
    exec catalog_server -p 1024 -l 100 -T 3
end script

• Start the catalog server and work_queue_pool

sudo start work_queue_pool
sudo start catalog_server

Install the Job Engine (Yerba)

Download and install the latest Yerba package from https://github.com/LyonsLab/Yerba/archive/v0.3.4.tar.gz

For more specific details on Yerba visit https://github.com/LyonsLab/Yerba/

The default installation path for Yerba will be in /opt/Yerba. If another path is chosen update the configuration files to match.

• Copy and the configuration file to /etc/yerba/yerba.cfg

[DEFAULT]
debug = True
access-log = /opt/Yerba/log/access.log
yerba-log = /opt/Yerba/log/yerbad.log

[yerba-log]
logging = /etc/yerba/logging.conf

[access-log]
logging = /etc/yerba/access.conf

[yerba]
port = 5151
level = DEBUG

[workqueue]
catalog_server = localhost
catalog_port = 1024
project = coge-main
log = /var/log/workqueue.log
port = -1
password = /etc/yerba/workqueue_pass
debug = True

[db]
path = /opt/Yerba/workflows.db
start_index = 100

• Copy the upstart file to /etc/upstart/yerba.conf

# /etc/init/yerba.conf

description "Yerba server daemon"
author "Evan Briones"

start on (local-filesystems and net-device-up IFACE=eth0)

stop on shutdown

respawn

pre-start script
    LOG_DIR=/opt/Yerba/log
    LOG_FILE=$LOG_DIR/debug.log
    [ -d "$LOG_DIR" ] || mkdir -m777 -p $LOG_DIR
#    [ -f "$LOG_FILE" ] || rm -f $LOG_FILE
end script

script
    export YERBA_ROOT=/opt/Yerba
    export PYTHONPATH="/usr/local/lib/python2.7/site-packages:$YERBA_ROOT"
    exec start-stop-daemon -c www-data -g www-data --start \
        --iosched real-time --nicelevel -19 \
        --exec $YERBA_ROOT/bin/yerbad -- >> $YERBA_ROOT/log/debug.log 2>&1
end script

post-start script
    echo Restart on: `hostname -A` | mail -s "UPSTART: Yerba was started" coge.genome@gmail.com
end script

• Initialize and start the job engine

/opt/Yerba/bin/yerbad --setup
sudo chown www-data:www-data /opt/Yerba/workflows.db
sudo start yerba

Troubleshooting

Visualization in GEvo does not work

This relies on a system known as Gobe. Check the following things:

• Apache configuration for gobe
• Make sure the Python Web module is installed: sudo aptitude install python-webpy
• Check to see if paths hard-coded into gobe/flash/service.wsgi need to be updated
  • NOTE: Not sure if this is required

Working on an Atmosphere Virtual Machine

Click here for instructions on dealing with issues that occur specifically with Atmosphere Virtual machines.