FRAGTE

FRAGTE performs fragment-based tetranucleotide frequency sieving to pre-select closely related prokaryotic genomes and reduce the computational burden of downstream whole-genome alignment and calculation for species delineation.

Key Features:

• Fragment-based sieving: Uses genome fragments rather than whole genomes for the sieving process, making sieving completeness-independent.
• Tetranucleotide frequency correlation: Employs fragment-based tetranucleotide frequency correlation to identify closely related genome pairs prior to alignment.
• Two-step strategy: Implements a sieving step to pre-select candidate pairs followed by alignment and calculation on the reduced set.
• Completeness independence: Avoids the >40% genome completeness requirement reported for genome-wide TETRA, enabling application to incomplete genomes.
• High sensitivity and specificity: Achieves approximately 100% sensitivity and high specificity across simulated genomes, real genomes, and metagenome-assembled genomes.
• Reduced downstream computation: Substantially lowers the number of genomes requiring alignment and calculation after sieving.
• Lower sieving cost: Decreases the computational cost associated with the sieving step itself.

Scientific Applications:

• Prokaryotic species delineation: Pre-selects candidate genome pairs for genome-wide approaches to species delineation in prokaryotes.
• Taxonomic studies: Streamlines computational demands in microbial genomics and taxonomic analyses that rely on whole-genome comparisons.
• Incomplete genomes and MAGs: Enables sensitive and specific sieving when working with incomplete genomes and metagenome-assembled genomes.

Methodology:

FRAGTE applies fragment-based tetranucleotide frequency correlation to sieve closely related genomes and then performs alignment and calculation on the reduced set of genome pairs; this two-step approach replaces genome-wide tetranucleotide correlation (e.g., TETRA) for the sieving step.