INTREPED

INTREPED predicts and annotates DNA repair proteins in newly sequenced genomes by integrating sequence, secondary-structure, and homology information to identify and classify DNA repair mechanisms.

Key Features:

• Homology-Based Detection: Uses homology information as a foundational method to identify DNA repair proteins.
• Machine Learning Integration: Employs Support Vector Machines (SVM) on feature sets derived from primary sequence polypeptide frequency, secondary structure, and homology data to classify DNA repair proteins and minimize false positives.
• Structural Analysis: Incorporates secondary structure information alongside primary sequence data to enhance protein identification accuracy.
• Clustering Techniques: Applies clustering methods to filter data and improve performance of machine-learning models that incorporate homology information.
• Genome Scanning and Annotation: Scans vertebrate genomes to detect novel DNA repair proteins, analyze correlations between genome size and repair protein transcript counts, and cluster organisms' repair abilities in an evolutionarily consistent manner.

Scientific Applications:

• Genomic Research: Enables comparative analyses of DNA repair mechanisms across organisms through genome-wide detection and annotation.
• Evolutionary Biology: Supports evolutionary clustering of repair capabilities and investigation of conservation of DNA repair pathways.
• Functional Genomics: Identifies putative, functionally unconfirmed proteins likely involved in DNA repair for experimental validation and functional characterization.

Methodology:

Combines sequence, secondary-structure, and homology data and applies Support Vector Machines (SVM) and clustering methods to predict DNA repair proteins and performs genome-scale scans of vertebrate genomes to assess correlations between genome size and repair gene counts.