GeneGrouper

GeneGrouper identifies and clusters gene clusters across diverse genomic datasets to enable population-level analysis of genetic variation and the detection of conserved and variable gene architectures.

Key Features:

• Density-based clustering: Employs a density-based clustering method to group gene clusters based on similarity.
• Binning by similarity: Categorizes gene clusters into bins according to their pairwise similarity.
• Cross-taxa detection: Detects gene clusters across phylogenetically proximate and distant taxa.
• Sensitivity to small differences: Identifies conserved gene clusters that differ by only a few genes, including cases resulting in pseudogenes or regulatory interruptions.
• Distant homologs and mobile elements: Clusters both distant homologous gene clusters and variable gene clusters found in integrative and conjugative elements.
• Population-level characterization: Provides population-level characterization of genetic variation within similar gene clusters.
• Benchmark performance: Demonstrated high recall and precision in benchmarks on a mixed-taxa dataset of 435 genomes.
• Specific cluster identification: Detected the 23-gene Salmonella enterica LT2 Pdu gene cluster and the four-gene Pseudomonas aeruginosa PAO1 Mex gene cluster in benchmark datasets.
• Discovery and experimental linkage: Uncovered a frequently occurring pduN pseudogene in 1130 S. enterica genomes and linked frameshift disruption of pduN to impaired microcompartment formation in vivo.

Scientific Applications:

• Comparative genomics: Enables identification and comparison of gene cluster conservation and variation across genomes.
• Pseudogene and regulatory variation detection: Facilitates detection of pseudogenes and regulatory disruptions within conserved clusters.
• Evolutionary and population studies: Supports population-level studies of genetic diversity and evolutionary dynamics of gene clusters.
• Study of mobile genetic elements: Enables analysis of variable gene clusters in integrative and conjugative elements.
• Phenotype association: Assists in linking cluster variation to phenotypic effects, as exemplified by pduN frameshift effects on microcompartment formation.

Methodology:

Applies a density-based clustering algorithm to categorize gene clusters into similarity-based bins.