ILM

ILM predicts RNA secondary structures including pseudoknots by applying an iterated loop matching algorithm that integrates thermodynamic and comparative sequence information.

Key Features:

• Iterated Loop Matching algorithm: Core algorithm that enhances prediction accuracy compared to traditional methods such as maximum weighted matching (MWM).
• Maximum Weighted Matching (MWM) support: Provides an alternative algorithm capable of detecting pseudoknots and tertiary base-pairing interactions.
• Integration of information types: Supports thermodynamic information, comparative sequence information, or a combination of both for structure prediction.
• Input compatibility: Handles individual RNA sequences and aligned sequences from multiple homologous RNA families.
• Prediction accuracy: Reports over 90% correct base-pair identification in short sequences and approximately 80% overall accuracy.
• Pseudoknot performance: Predicts nearly all pseudoknot structures while minimizing false positives in sequences without pseudoknots.
• Output formats: Produces predicted RNA secondary structures in formats compatible with existing visualization tools.
• Computational efficiency: Demonstrates higher efficiency and lower resource requirements than methods such as PKNOTS while maintaining high accuracy.
• Implementation: Implemented in ANSI C.

Scientific Applications:

• RNA secondary structure prediction with pseudoknots: Modeling and identification of base pairs and pseudoknots in individual RNAs and aligned homologs.
• Comparative analysis of homologous RNA families: Leveraging multiple-sequence alignments to improve structure inference across related RNAs.
• Studies of RNA folding, function, and evolution: Using predicted structures to investigate folding patterns, functional motifs, and evolutionary conservation.
• Large-scale or resource-constrained analyses: Application where computational efficiency relative to tools like PKNOTS is a priority.

Methodology:

Implements an iterated loop matching algorithm and optionally maximum weighted matching (MWM), integrating thermodynamic and comparative information; implemented in ANSI C.