Prediction of nucleosome occupancy in Saccharomyces cerevisiae using position-correlation scoring function

Area of Science:

• Genomics
• Molecular Biology
• Bioinformatics

Background:

• Nucleosome organization and positioning are crucial for gene regulation and expression.
• Understanding these mechanisms requires detailed genomic analysis.

Purpose of the Study:

• To develop and validate an algorithm for distinguishing nucleosome and linker DNA sequences based on 4-mer frequency bias.
• To predict nucleosome occupancy across the S. cerevisiae genome using the developed algorithm.

Main Methods:

• Statistical analysis of 4-mer frequency bias in S. cerevisiae nucleosome and linker sequences.
• Development of a position-correlation scoring function algorithm.
• Five-fold cross-validation to assess algorithm performance (mean AUC = 0.981).
• Genome-wide prediction of nucleosome occupancy and comparison with experimental maps.

Main Results:

• The algorithm effectively distinguishes nucleosome from linker sequences.
• High correlation coefficients were obtained between predicted and experimental nucleosome positioning maps.
• Distinct nucleosome-depleted regions near regulatory sites were confirmed.

Conclusions:

• Intrinsic DNA sequence preferences in linker regions significantly influence nucleosome occupancy.
• The developed algorithm provides a reliable tool for predicting nucleosome positioning and understanding gene regulation.