A novel sensitive method for the detection of user-defined compositional bias in biological sequences

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Biological sequences frequently contain compositionally biased segments where specific residue types are overrepresented.
• The function, evolution, and significance of these segments are often overlooked in standard sequence analysis.
• However, biased segments with similar chemical properties in amino acids are linked to molecular functions and human diseases.

Purpose of the Study:

• To develop a sensitive method for detecting compositionally biased segments based on user-defined residue types.
• To enable large-scale analysis of the functional implications and evolutionary conservation of these segments.

Main Methods:

• Introduction of BIAS, a novel method utilizing discrete scan statistics for bias detection.
• BIAS computes analytical significance estimates, correcting for multiple tests and accounting for global compositional bias.
• Benchmarking BIAS against existing methods like SEG, SAPS, and CAST.

Main Results:

• BIAS demonstrates high sensitivity in detecting user-specified compositionally biased segments.
• The method accurately corrects for multiple testing and considers global compositional bias.
• BIAS analysis reveals significant associations between specific protein functional groups and particular types of biased segments.

Conclusions:

• BIAS provides a sensitive and accurate tool for analyzing compositionally biased segments in biological sequences.
• The findings highlight the functional relevance of these segments, linking them to protein function and potential disease involvement.
• This method facilitates deeper understanding of the role of compositional bias in molecular evolution and function.