Carlos Masdeu Avila & Francesc Castro Giner

2.-Biological Problem
3.-The Program


For understand the transcriptional regulatory mechanism that operate in genes, it's important to realize a mapping of transcription factor binding sites in genes. Now, there are many of these site experimentally identified that can be used to perform a computational pattern-based searches. Motifs are a short and variable regions that will produce over-predictions problems.
We have developed a new program (in Perl language) for deteccion of promoter motif in a DNA sequences. Motifs,are descripted in a Position Weight Matrix (PWM) from TRANSFAC database. Position Weigth Matrix is a motif descriptor. It attempts to capture the intrinsic variability characteristic of sequence patterns. It is usually derived from a set of aligned sequences functionally related.

2.-Biological Problem

The regulation of gene expresion in eukariotes is a complex process that is difficult to understand due to variability of the mechanism involved in and the great number of different actors playing some minor o major role.
How is gene expression regulated?There are several methods used by eukaryotes.

Protein-encoding genes have
Eukariotic Promoter

Transcription factors (TF)are proteins that bind to promoter of a gene. These proteins either inhibit or assist RNA polymerase in initiation and maintenance of transcription. DNA binding domain: amino acids in the protein that recognize specific bases near the start of transcription.
TF are attrached to the promoter region by very specific motifs imprinted in the DNA called binding sites. Tf are usually arranged along the promoter region following very restricted rules such as minimum/maximum distance or neighbourhood constrains which must therefore be verified when predictin putative DNA binding sites. The problem of finding promoters regions is extremely difficult because there are hundreds of differents TFs, their hability to connect to more than one different motif, the short lenght of motif (5-15 pb) and the poor knowledge about original interactions.

3.- The Program

Download it!

Promotif will detect promoter motif in a DNA sequence using a PWM
Inputs: Tranfac matrix and fasta sequence
ProMotif compares input PWM and input sequence. It compares every subsequence (with a position numbers like matrix- positions number)of input sequence, whit PWM, and calculate a score.
We can choose the type of PWM that ProMotif use to compute the scores:

The program filter the scores, and make a output with the sequences which score is higher than a treshold. The default treshold is 0'85.This program is capable of read and scan more than one sequence and more than one matrix

matrix_file.mat: should be in Transfac format.
sequence_file.txt: should be in Fasta format.

Basic program ejecution
$./perl matrix_file.mat sequence_file.txt


	-c_x.x	:Cutoff value
	-v	:Show information of the processing
	-n_x	:Chose type of matrix-Relative Matrix
			Default matrix: relative	
			x=1 -Log-likelihood matrix(whit a priory frequency-0'25-)
			x=2 -Log-likelihood matrix(frequency in the aligned region)
	-m	:Show information about the matrices
	-s	:Show information about DNA sequences (C+G content, lenght)
	-o	:Makes a HTML Output
	-h	:help

Execution example:
$./perl -svc 0.55 matrix_file.mat sequence_file.txt

4.- Test

With the finality of test our program we have designed an exercise. We searched a HNF1alpha motif in a problem sequences. We obtained he HNF1alpha PWM in TRANSFAC database. To compare the results ofproblem sequences, we have maked a negative control using a aleatory sequences obtained whith a program(AleatorySequences) We run ProMotif with these input files:

We test ProMotif whith diferent tresholds. This table is a results summary:

Output for problem sequences Output for aleatory sequences
0.95 1 0.95 0
0.85 10 0.85 0
0.80 33 0.80 5

We compare the problem sequences with aleatory sequences in order to show that the reults of ProMotif aren't due to fate.
If we compare the outputs for the two cases, and each treshold, we can observe a major occurrence number in outputs of problem sequences. For example, in study of aleatory sequences with a tresold of 0.95 and 0'85, we obtained no occurrences, whereas that in problem sequences we have obtained some occurrences.

In conclusion, we can say that ProMotif search specífic motif, and his reults aren't due to fate.

5.- Bibliography
-MatInd and MatInspector: new fast and versatile tools for detection of consensus matches in nucleotide sequence data Quandt K, Frech K, Karas H, Wingender E, Werner T. Nucleic Acids research, 1995, Vol.23, No.23

Comments, suggestions, death threats and proposals of marriage to:
Carlos Masdeu & Francesc Castro

#Thanks to E.Blanco for his contributions.