GeneStitch

GeneStitch reconstructs full-length genes from fragmented de Bruijn graph-based metagenomic assemblies by matching contigs to reference gene sets with a network matching algorithm to resolve homologous-gene-induced repeats.

Key Features:

• Network matching algorithm: Aligns contigs within a de Bruijn graph to reference genes to identify "gene paths" and connect fragmented gene sequences into longer assemblies.
• De Bruijn graph integration: Operates on de Bruijn graph-based assemblies to traverse assembly graph structures and link contigs that represent parts of the same gene.
• Reference gene set integration: Uses homologous and distantly related reference genes as guides to detect contig sequences with high similarity and enable reconstruction of more complete genes.

Scientific Applications:

• Metagenomic gene assembly improvement: Increases completeness of gene assemblies from complex metagenomic datasets that contain homologous genes among closely related species.
• Functional annotation enhancement: Produces more complete gene sequences to improve downstream functional annotation and gene function prediction.
• Microbial community and ecosystem analysis: Aids interpretation of microbial community gene content and interactions in environmental and clinical metagenomes.

Methodology:

Matches contigs from de Bruijn graph assemblies to reference genes using a network matching algorithm to identify optimal "gene paths" (sequences of contigs with highest similarity to known genes) and connects these fragments into more complete genes.