RF_DYMHC

RF_DYMHC predicts recombination hot and cold spots in the yeast genome to identify regions with differential meiotic double-strand break (DSB) frequency.

Key Features:

• Random Forest Prediction Model: Implements a random forest (RF) algorithm to analyze DNA sequence data for predicting recombination hotspots and coldspots.
• Performance Metrics: Reports an overall accuracy of 82.05% and a Matthew's correlation coefficient (MCC) of 0.638 for hotspot versus coldspot classification.
• Comparison with SVM: Demonstrates superior sensitivity and specificity relative to a support vector machine (SVM) benchmark.
• Prediction Parameters: Incorporates RI-value and non-overlapping window scan size parameters for sequence scanning and prediction.
• Visualization of Results: Marks predicted hot and cold spots on genomic sequences for visual interpretation of predicted regions.

Scientific Applications:

• Meiotic recombination mapping: Mapping hotspots and coldspots to elucidate mechanisms of meiotic DSB initiation.
• Regulatory element identification: Identifying potential regulatory elements involved in meiotic DSB formation and repair.
• Genome stability and diversity studies: Investigating determinants of genetic diversity and genome stability in eukaryotic organisms.

Methodology:

Applies a random forest algorithm to genomic DNA sequence data and uses out-of-bag (OOB) estimation for model performance validation.