RNAstrand

RNAstrand predicts the reading direction of structured RNAs from multiple sequence alignments using machine learning and thermodynamic features.

Key Features:

• Support Vector Machine Classification: Applies a support vector machine (SVM) to classify the correct strand orientation of structured RNAs based on alignment-derived features.
• Energy and Evolutionary Feature Integration: Utilizes folding energy differences between RNA sequences and their reverse complements together with substitution patterns from multiple sequence alignments to improve strand discrimination accuracy.

Scientific Applications:

• ncRNA Functional Annotation: Determines correct reading direction to support accurate annotation and functional characterization of non-coding RNA genes.

Methodology:

RNAstrand analyzes folding energies of aligned RNA sequences and their reverse complements, extracts evolutionary substitution patterns, and integrates these features into an SVM model to classify strand orientation with enhanced accuracy.