MycoAlign Installation Procedure

From GSAF

How to install MycoAlign.

The goal is to be able to hand a vanilla Linux machine and this wiki page to an administrator, and they could install MycoAlign from scratch without knowing anything about what MycoAlign is or what it does.

[edit] Prerequisites

Linux
MySQL
Perl, C compiler, Python TODO: details?
Apache, FastCGI? TODO: details?

[edit] Installation

adduser i3bio       # Production
adduser i3biodev    # Developers / QA
cd /home
svn checkout https://hive.gds.unomaha.edu/svn/i3bio/tags/current i3bio
svn checkout https://hive.gds.unomaha.edu/svn/i3bio/trunk        i3biodev

# Compile "compare", a C program:
cd i3bio/public_html/db/compare
make

# Build the database
cd /home/i3bio/
... create the MySQL database "i3bio"       TODO: Procedure?
... create the MySQL database "i3biodev"    TODO: Procedure?
... Security? Usernames and passwords?      TODO: Procedure?
mysql i3bio    < public_html/db/scripts/schema.sql
mysql i3biodev < public_html/db/scripts/schema.sql

# Configure Apache
... Are the Apache configs in SVN? Documentation? FastCGI?  TODO: How?

vi public_html/db/config.pl
   # Change settings to production database (username, passwords, etc. Details here.
   # (It's the i3biodev config that is committed to SVN, not the production config.)

[edit] Create /tmp/Actinobacteria database for subsequent searching

The CLAB server has most of the current GenBank archive on it. So let's pop over there and grab a current dump of all Actinobacteria sequences. We'll convert those sequences to FASTA, compress the file, move it to the i3bio server, and prepare for BLAST queries.

ssh clab.ist.unomaha.edu
cd /sequences/genbank/
seqget.pl --organism Actinobacteria 1>/tmp/A.genbank
genbank2fasta.pl /tmp/A.genbank /tmp/A.fasta
bzip2 /tmp/A.fasta
scp /tmp/A.fasta.bz2 i3bio.gds.unomaha.edu:/tmp/

ssh i3bio.gds.unomaha.edu
mkdir /tmp/Actinobacteria
cd /tmp/Actinobacteria
mv ../A.fasta.bz2 ./
bzip2 -d A.fasta.bz2
formatdb -i A.fasta -o T -p F

Now your directory should look like this:

jhannah@i3bio:/tmp/Actinobacteria2$ ls -al
total 332144
drwxr-xr-x  2 jhannah jhannah      4096 2008-01-24 18:45 .
drwxrwxrwt 47 root    root         8192 2008-01-24 18:46 ..
-rw-r--r--  1 jhannah jhannah 268569337 2008-01-24 18:41 A.fasta
-rw-r--r--  1 jhannah jhannah   4122667 2008-01-24 18:45 A.fasta.nhr
-rw-r--r--  1 jhannah jhannah    408580 2008-01-24 18:45 A.fasta.nin
-rw-r--r--  1 jhannah jhannah   1123780 2008-01-24 18:45 A.fasta.nsd
-rw-r--r--  1 jhannah jhannah     26104 2008-01-24 18:45 A.fasta.nsi
-rw-r--r--  1 jhannah jhannah  65485509 2008-01-24 18:45 A.fasta.nsq
-rw-r--r--  1 jhannah jhannah       600 2008-01-24 18:45 formatdb.log

So as long as these files look big and fat: nhr nin nsd nsi nsq, you're done w/ A.fasta. So let's remove it since it's huge and we don't need it any more.