DBPMod

DBPMod predicts species-specific DNA-binding proteins (DBPs) using supervised machine learning to enable accurate identification across model organisms for studies of gene expression, DNA replication, recombination, and repair.

Key Features:

• Target entity: Focuses on DNA-binding proteins (DBPs) relevant to gene expression, DNA replication, recombination, and repair.
• Species-specific prediction: Employs a supervised learning model tailored to predict DBPs within distinct species.
• Algorithms: Integrates both shallow and deep learning algorithms for model construction and comparison.
• Shallow learning performance: Reports superior accuracy from shallow learning models compared with deep learning models for the tested tasks.
• Feature representation: Uses evolutionary features rather than sequence-derived features for prediction.
• Benchmark organisms: Evaluated across Caenorhabditis elegans, Drosophila melanogaster, Escherichia coli, Homo sapiens, and Mus musculus.
• Validation strategies: Performance assessed using five-fold cross-validation and independent test set analyses.
• Performance metrics: Reports auROC values of approximately 89–92% and auPRC values of 89–95%.
• Comparative benchmarking: Compared against twelve state-of-the-art computational approaches and reported consistent outperformance across tested organisms.

Scientific Applications:

• Species-specific DBP discovery: Identification of DNA-binding proteins in Caenorhabditis elegans, Drosophila melanogaster, Escherichia coli, Homo sapiens, and Mus musculus.
• Functional genomics support: Supports studies of molecular mechanisms involving gene expression, DNA replication, recombination, and repair by providing candidate DBPs.
• Computational benchmarking: Provides a comparative framework for evaluating DBP prediction performance against existing methods.

Methodology:

Uses supervised learning models (shallow and deep learning) trained on evolutionary features, with evaluation by five-fold cross-validation and independent test set analyses reporting auROC and auPRC metrics.