GeneFizz

GeneFizz leverages helix-to-coil transitions to identify genes in DNA sequences by comparing physics-based structural segmentation (helix vs coil) with coding/non-coding annotations.

Key Features:

• Physics-based structural segmentation: Derives helical (helix) and non-helical (coil) domains from the classical helix-coil model of DNA structure.
• Coding/non-coding segmentation comparison: Compares physics-based structural segmentation with coding/non-coding segmentation aligned to simple or split genes and exons.
• Ab initio gene identification: Performs physics-based, ab initio identification of novel genes from sequence-derived structural properties without relying on homology.
• Split-gene and exon sensitivity: Addresses challenges in predicting complex split genes and exon structures in eukaryotic genomes.
• Segmentation coincidence analysis: Reveals varying degrees of coincidence between structural and annotation-based segmentations across different genomes.
• Evolutionary signal exploration: Uses discrepancies between physics-based and annotation-based segmentations to explore potential evolutionary signals.
• Empirical validation: Has been applied to the Plasmodium falciparum genome and identified previously unrecognized genes that lack homology to known sequences.
• Integration with genomic annotations: Integrates structural DNA properties with traditional genomic annotations to support gene discovery and comparative analyses.

Scientific Applications:

• Ab initio gene discovery in eukaryotes: Identification of novel genes, including complex split genes, in eukaryotic genomes using structural DNA features.
• Discovery of non-homologous genes: Detection of genes lacking homology to known sequences through physics-based analysis.
• Comparative genomics and evolutionary analysis: Exploration of evolutionary signals by comparing structural segmentation with annotation-based segmentation across genomes.
• Genome-specific validation: Application to and validation on the Plasmodium falciparum genome for discovery of previously unrecognized genes.

Methodology:

Applies the classical helix-coil model to derive helix/coil structural segmentation and employs the Physics-Based Gene Identification (PBGI) approach to compare these segmentations with coding/non-coding annotations for ab initio gene identification.