GIIRA

GIIRA predicts genes in prokaryotic and eukaryotic genomes by integrating RNA-Seq data and reallocating ambiguously mapped reads to improve identification of expressed coding regions.

Key Features:

• RNA-Seq integration: Uses high-throughput RNA-Seq data to inform gene prediction.
• Ambiguous read handling: Retains and explicitly handles ambiguously mapped reads rather than discarding them.
• Candidate region extraction: Extracts candidate genomic regions based on a sufficient number of RNA-Seq read mappings.
• Maximum-flow reassignment: Employs a maximum-flow algorithm to reassign ambiguous reads to their most probable origin.
• Cross-domain support: Applicable to both prokaryotic and eukaryotic genomes for coding region identification.
• Validation: Evaluated on simulated and real datasets and reported improved performance compared to existing methods that incorporate RNA-Seq.
• Implementation: Implemented in Java.

Scientific Applications:

• Gene prediction: Identification of expressed coding regions from RNA-Seq data in prokaryotic and eukaryotic genomes.
• Recovery of genes with ambiguous support: Rescue of genes predominantly supported by ambiguously mapped reads.
• Method benchmarking: Comparative evaluation of gene-finding accuracy on simulated and real RNA-Seq datasets.

Methodology:

Extracts candidate regions based on a threshold of RNA-Seq read mappings and employs a maximum-flow algorithm to reassign ambiguously mapped reads to their most probable origin; implemented in Java.