README for RPMCMC motif sampler

Current version(0.0) of RPMCMC has some limitations as follows:
	maximum number of input sequences is 100000
	maximum length of input sequences is 3000
	maximum iteration is 10000

##Installation
GCC version 4.4 or above is required since this software uses OpenMP v3.0.
For Windows, installing linux in a virtual machine might be preferable. 

 % cd [download directory]
 % tar xvfz rpmcmc-xxx.tar.gz
 % cd ./rpmcmc-xxx/src
 % make
 % make install

##stack size
Before executing the application, it is necessary to increase the stack size as following command:
 % ulimit -s unlimited

##Start searching motifs
 % rpmcmc-xxx/bin/multi_motif_sampler -d [input fasta file] -od [output directory] 

##Option
 % rpmcmc-xxx/bin/multi_motif_sampler [-key value]  
 	od:output directory path:
	d:input fasta file path:
	ps:input position specific prior file path:(optional; uniform prior by default)
	p:order of markov background model(0<=p<=3)[int] (optional; 0 by default)
	ks:motif length min(ks<=kl<=30)[int] (optional; 6 by default)
	kl:motif length max(ks<=kl<=30)[int] (optional; 14 by default)
	n:the number of replicas(1<=n<=100)[int] (optional; 20 by default)
	r:strength of repulsive force(0<=r<=100)[float] (optional; 10.0 by default)
	i:the number of iteration(u<=i<=10000)[int] (optional; 520 by default)
	u:burnin step(0<=u<i)[int] (optional; 20 by default)
	b:prior of motif presence(0<b<=1)[float] (optional; 0.76 by default)
	g:penalty on different length(g>=0)[float] (optional; 0.3 by default)
	l:threshold for clustering(0<=l<=1)[float] (optional; 0.3 by default)

##Sample data
 test_sequence.fasta is a sample dataset placed in src directory (rpmcmc-xxx/src/test_sequence.fasta). 

##Example
 % rpmcmc-xxx/bin/multi_motif_sampler -d rpmcmc-xxx/src/test_sequence.fasta -od ./ -n 10 -i 120

 RPMCMC will start searching motifs in the sample dataset with 10 replicas and 120 iterations.
 Output files will be placed in the current directory (option key -od). 

##Output
filename:pfm
This file shows the Position Frequency Matrix(PFM).

*Example
Motif1          ## motif index
A C G T         ## order of letters
335 500 73 1976 ## frequency of each letters
21 29 63 2771 
16 7 26 2835 
1 2801 58 24 
31 2067 24 762 
37 430 16 2401 

Motif2          ## next motif
A C G T
121 2663 149 0 
2620 294 4 15 
8 2488 47 390 
41 188 2591 113 
103 190 47 2593 
54 15 2863 1 

filename:locate
This file shows where motifs are embedded.

*Example
Motif1
seq_no start end direction motifSeq
1 1 6 - TTTCCC
2 378 383 - GTTCTT
3 53 58 - TTTCCT
4 34 39 - TTTCCC
6 84 89 - ATTCCT
7 389 394 - CTTCCT
8 364 369 - CTTCTT
9 73 78 + ATTCCT
10 233 238 + TTTCTT
11 408 413 - TTTCCT
12 196 201 + CTTCCT
13 455 460 - CTTCCT
14 82 87 - TTTCCT
15 111 116 - TTTCCT
16 349 354 + TTTCTT
17 64 69 - TTTCCT

First column is the index number of sequences in fasta sequence set.
Second and third columns are the start and the end positions, respectively.
Fourth column is the direction ('+' and '-' mean that they locate from 5' to 3' and 3' to 5', respectively).
Fifth column shows each sequence.

###contact
ikebata.hisaki@ism.ac.jp
Feel free to asking if you do not understand how to use RPMCMC.