Proteinortho

Proteinortho identifies orthologous genes across multiple species to support comparative genomics, phylogenomics, and evolutionary studies.

Key Features:

• Algorithmic Efficiency: Implements a graph-based detection method using a reciprocal best alignment heuristic optimized by a pseudo-reciprocal best alignment strategy, reducing the number of necessary sequence comparisons by half and accelerating initial protein similarity identification.
• Scalability and Memory Efficiency: Proteinortho6 reduces processing time by an order of magnitude compared to its predecessor, maintains a small memory footprint, and employs a reworked clustering algorithm to efficiently decompose large clusters.
• Distributed Computing: Leverages distributed computing on multi-core systems to compute genome-wide orthology relations without relying solely on pre-existing database entries.
• Orthology Analysis for Large Datasets: Capable of processing very large datasets, as demonstrated by analysis of 717 eubacterial genomes from NCBI to identify proteins present in the majority of bacterial proteomes.
• Enhanced Orthology Detection with PoFF: The PoFF extension integrates clustering, sequence similarity, and synteny analysis and employs the FFAdj-MCS heuristic to assess pairwise gene order using adjacencies, supporting multiple linear chromosomes and detecting duplicated regions to reduce false positives.

Scientific Applications:

• Comparative Genomics: Identification of orthologous genes to compare genetic content across species.
• Functional Genomics: Orthology assignments for predicting gene function by leveraging known functions of orthologs in other organisms.
• Phylogenomics and Evolutionary Studies: Support for reconstruction of phylogenetic trees and study of sequence evolution patterns to investigate species divergence and adaptation.

Methodology:

Graph-based detection using a reciprocal best alignment heuristic and a pseudo-reciprocal best alignment strategy (reducing comparisons by half); a reworked clustering algorithm for decomposing large clusters; distributed multi-core computing; and, in PoFF, integration of clustering, sequence similarity and synteny analysis using the FFAdj-MCS heuristic to assess pairwise gene order via adjacencies, supporting multiple linear chromosomes and detecting duplicated regions.