Computational discovery of feature patterns in nucleosomal DNA sequences

Area of Science:

• Molecular Biology
• Genomics
• Bioinformatics

Background:

• Nucleosome formation mechanisms are crucial for understanding gene regulation.
• Previous research suggested multiple features influence nucleosome assembly.

Purpose of the Study:

• To develop a computational method for identifying key features and patterns involved in nucleosome formation.
• To investigate the combined effect of various features on nucleosome occupancy.

Main Methods:

• Compiled 779 potential features related to DNA sequence and structure.
• Developed the Finding Features for Nucleosomes (FFN) algorithm for pattern discovery and scoring.
• Applied FFN to genome-wide nucleosome occupancy data from yeast and human.

Main Results:

• Identified statistically significant feature patterns influencing nucleosome formation, many conserved across yeast and human.
• Demonstrated the importance of both sequence and structural features for predicting nucleosome occupancy.
• Showed that variations within feature combinations impact predictive power.

Conclusions:

• FFN effectively identifies patterns governing nucleosome formation.
• Conserved feature patterns highlight fundamental principles of nucleosome organization across species.
• The identified patterns can aid in predicting nucleosomal sequences and understanding gene regulation.