p6mA

p6mA predicts N6-methyladenine (6mA) sites in DNA sequences using sequence-derived features and machine learning to identify methylated loci across multiple species.

Key Features:

• Sequence-Based Prediction: Utilizes sequence-derived features to predict 6mA methylation status from DNA sequences, reducing reliance on wet‑lab assays.
• Feature Integration: Integrates physicochemical properties, position-specific triple-nucleotide propensity (PSTNP), and electron‑ion interaction pseudopotential (EIIP) with feature selection via maximum relevance maximum distance (MRMD).
• Machine Learning Algorithm: Employs Extreme Gradient Boosting (XGBoost) for predictive modeling of 6mA sites.
• Model Versatility: Provides six built-in models tailored to different species including Oryza sativa, Caenorhabditis elegans, Drosophila melanogaster, and Homo sapiens, and includes Compre and Compre2 models trained on comprehensive and non-redundant comprehensive datasets, respectively.

Scientific Applications:

• Epigenetic analysis: Predicts distribution patterns of 6mA across organisms to support studies of DNA methylation related to gene regulation, development, and disease mechanisms.

Methodology:

Extracts sequence-based features (physicochemical properties, PSTNP, EIIP), applies MRMD for feature selection, and trains XGBoost models on multi-species datasets including comprehensive (Compre) and non-redundant comprehensive (Compre2) datasets.