RescueNet

RescueNet employs the Self-Organizing Map (SOM) algorithm to identify multiple gene models from Relative Synonymous Codon Usage (RSCU) patterns and improve prediction of protein-coding regions with atypical sequence compositions within genomes.

Key Features:

• Self-Organizing Map Algorithm: RescueNet employs the Self-Organizing Map algorithm to discover and organize distinct codon usage patterns without prior specification of cluster number.
• Multiple Gene Models: RescueNet identifies and represents diverse intra-genomic gene models to capture compositional variation in protein-coding regions.
• Relative Synonymous Codon Usage (RSCU) analysis: RescueNet uses RSCU metrics to distinguish coding regions from non-coding sequences based on codon usage biases.
• Complementary Use: RescueNet can detect genes overlooked by single-model gene prediction methods, providing complementary predictions for genome annotation.

Scientific Applications:

• Genome annotation: RescueNet can enhance genome annotation by detecting genes with non-standard codon usage patterns.
• Gene-finding method development: RescueNet can be used to develop and evaluate gene-finding techniques that accommodate intra-genomic compositional variation.
• Evolutionary and functional genomics: RescueNet supports studies of codon usage, genetic functions, and evolutionary biology by revealing diverse gene models across genomes.

Methodology:

RescueNet applies a Self-Organizing Map algorithm to RSCU-derived codon usage data to identify multiple intra-genomic gene models and discriminate coding from non-coding sequences.