psilc: A Java repository from lachlancoin

REQUIREMENTS: java 1.4.x or higher (run on java1.4.2_05, not tested on other versions)

See the examples/ directory with example commandlines in run.sh and runGraphical.sh for examples.

USAGE:

java -Xmx300m -jar psilc.jar --fasta seed.dna.fa

or (inputing an aligned rather than unaligned dna sequence)

java -Xmx300m -jar psilc.jar --align seed.dna.mfa

or for a graphical interface (mostly undocumented for the moment, but email me with questions)

java -Xmx 300m -cp psilc.jar lc1.treefam.TreeCurationTool

This will create an alignment called seed.dna.mfa (using muscle). You can supply your own
seed.dna.mfa if you want PSILC to use your own alignment (in FASTA or CLUSTAL format). Otherwise it will use MUSCLE.

It will also create a tree called seed.nhx. Again if you supply your own (in nh or nhx format), this will be used. Otherwise the
tree is created using phyml. If it can't find phyml, it will create a neighbour joining tree using ml distances under a WAG rate
matrix with 4 gamma rates.

With the following optional parameters
--color [pseudo|selection] whether to colour and label the branches according to pseudo gene or selection index

--seq < sequence identifier to score >

if this is not specified, all leaf nodes are scored as default. You must also specify --restrict 1

--repository <directory which contains the "Pfam_ls" file, and/or a directory called Pfam, containg Pfam HMMs >
if this is not specified, then the directory "." is used.
The "Pfam_ls" file is downloadable from
ftp://ftp.sanger.ac.uk/pub/databases/Pfam/Pfam_ls.gz
It is a flatfile of the hidden markov models used by Pfam. You can build your own using HMMER.

--bin <bin_directory1:bin_directory2>
all specified directories will be searched for the binaries "muscle", "phyml". If these binaries are in your CLASSPATH then you don't
need to specify this flag.

--graph [0|1] - whether or not to display a graphical view of the
results. If this flag is set to one, then a tree is displayed of the
alignment, and the user can (by clicking on the nodes) bring up a
graph of the position by position PSILC scores, as well as the
posterior probabilities of being in a pseudogene/selected/ normal
state.

--allNodes [0|1] - if this flag is set to one all nodes (including internal nodes) are scored.

--nucrm [HKY, GTR, TN] - the nucleotide rate matrix to use (default HKY)

--protrm [WAG, JTT] - the protein rate matrix to use (default WAG)

It may be necessary to enable java to access extra memory, by running, for instance

java -Xmx200m psilc.jar

During the first run, and Pfam_ls_idx file will be created to index
the Pfam_ls file. This step will be skipped on future runs, providing
the Pfam_ls_idx file is in the right directory.

Both alignment and fasta inputs must contain DNA sequences.

OUTPUT:

"PSILC_WAG_HKY" - this contains the PSILC results (assuming a WAG model of protein evolution and an HKY model of DNA evolution,
which can be changed using --protrm and --nucrm flags (see above).

Source:

The source is available on request, email lc1@sanger.ac.uk

lachlancoin/psilc