RecGroup program

This program is aimed on the construction of Recognition Groups which are used further by the <RGSiteScan> for obtaining the final result, that is, the revealing of potential binding sites in a DNA sequence. So this is the first program in a two-stage process of revealing of transcription factor binding sites.

Recognition Group (RG) represents a high homology set of oligonucleotides of equal lengths. Recognition Group retains most of the information and is, thus, well suited to describe the transcription factor binding sites.

Recognition Group is built on the basis of the sample of non-aligned DNA sequences of various lengths, which a priori contain the binding sites. There may be no information about the location and strand orientation of the sites contained in these sample sequences. This is the advantage of the algorithm used compared with the standard approaches, such as consensus method and nucleotide frequency matrix approaches.

Each oligonucleotide from RG is accomplished with the weight that reflects a proportion of sample sequences containing this oliginucleotide in an arbitrary orientation.

The site length l (i.e. oligonucleotide length) is an estimable parameter as well as the degree of homology T. So, one of the results provided by this program is a precisely defined site length. Both of these parameters are estimated by means of original two-parametrical optimization procedure adapted to the task pursued.

EXAMPLE

Input:

The expert set for NF-Y transcription factor.

*⁾ the EMBL-sequence HSGAMGLOA contains two different actual binding sites simultaneously

Output:

The recognition group consisting of 10 oligonuleotides l=7bp long for the description of NF-Y (CCAAT-binding factor) transcription factor binding site:

Oligonuleotide weight