[Pattern recognition in the computer analysis of nucleotide sequences].

Area of Science:

• Bioinformatics
• Computational Biology
• Molecular Biology

Background:

• Bacterial gene regulation relies on promoter sequences.
• Accurate identification of promoter elements is crucial for understanding gene expression.
• Escherichia coli serves as a model organism for studying prokaryotic gene regulation.

Purpose of the Study:

• To develop a computational method for identifying distinguishing features of Escherichia coli promoters.
• To create a recognition matrix for predicting promoter sequences.
• To analyze the occurrence and strength of predicted promoters.

Main Methods:

• Application of the "generalized portrait" algorithm from pattern recognition theory.
• Selection of significant sequence signs (features) for promoter identification.
• Multiple sequence alignment to refine promoter models.
• Calculation of a recognition vector (matrix) for promoter scoring.

Main Results:

• A distinguishing vector (recognition matrix) for Escherichia coli promoters was successfully generated.
• The algorithm effectively ranked promoters based on their known strength.
• Analysis confirmed the occurrence of predicted promoter sequences.

Conclusions:

• The "generalized portrait" algorithm is effective for identifying Escherichia coli promoter signals.
• The developed method facilitates the prediction and analysis of promoter elements.
• A software tool for promoter searching is available to researchers.