CoreCruncher

CoreCruncher constructs core genomes from large prokaryotic datasets by identifying genes shared across strains using a heuristic classification of orthologs versus paralogs and xenologs based on distributions of identity scores.

Key Features:

• Core genome construction: Efficiently and robustly constructs core genomes from large prokaryotic datasets.
• Heuristic approach: Avoids exhaustive pairwise genome comparisons through a heuristic classification strategy.
• Identity score distributions: Leverages distributions of identity scores to distinguish orthologs from paralogs and xenologs.
• Ortholog/paralog/xenolog classification: Classifies sequences into orthologs or paralogs/xenologs to inform core gene inference.
• Conservative specificity: Adopts a conservative strategy that minimizes false positives and can reduce sensitivity to paralogs and xenologs.
• Alignment tool compatibility: Uses Usearch or Blast for sequence alignment and can use muscle or mafft for specific functions.
• Implementation and dependencies: Implemented in Python 3.7, compatible with Python 2.7, and requires Numpy.
• Speed and robustness: Operates faster than methods requiring exhaustive pairwise comparisons while maintaining robustness in core genome construction.

Scientific Applications:

• Core genome inference: Constructs genes shared by all or nearly all strains within prokaryotic populations or species for downstream analyses.
• Comparative genomics: Identifies conserved genes across prokaryotic strains and species while minimizing false-positive shared-gene calls.
• Genetic conservation analyses: Supports studies of conserved gene content and genetic conservation across prokaryotic species.

Methodology:

Employs a heuristic that circumvents exhaustive pairwise genome comparisons, leverages distributions of identity scores to classify sequences as orthologs or paralogs/xenologs, performs sequence alignments with Usearch or Blast (with certain functions using muscle or mafft), and is implemented in Python 3.7 (compatible with Python 2.7) requiring Numpy.