Orthograph

Orthograph assigns orthologs and paralogs by mapping nucleotide and amino acid sequences to the globally best matching clusters of orthologous genes using a best reciprocal hit strategy with profile hidden Markov models (HMMs), enabling reference-based orthology prediction for genomic and transcriptomic (including RNA-seq) data.

Key Features:

• Best reciprocal hit strategy: Implements a reciprocal search approach to identify globally best matching orthologous clusters.
• Profile HMM mapping: Uses profile hidden Markov models (HMMs) to map nucleotide and amino acid sequences to orthologous gene clusters.
• Sequence type support: Handles both nucleotide (including coding sequences from RNA-seq) and amino acid sequences.
• Data source compatibility: Operates on genomic and transcriptomic data and is applicable to partial datasets such as de novo-assembled transcriptomes.
• Orthology and paralogy delineation: Differentiates orthologs and paralogs for comparative analyses and functional inference.
• Validation: Validated on de novo-sequenced and assembled transcript libraries from 24 apoid wasp species (Hymenoptera: Aculeata) and other published genomic datasets.

Scientific Applications:

• Comparative genomics: Assigns orthologs and paralogs to support cross-species evolutionary analyses.
• Transcriptome-based orthology: Enables reference-based orthology prediction for coding nucleotide sequences derived from RNA sequencing and de novo transcriptomes.
• Functional annotation transfer: Facilitates functional annotation across species by identifying reliable orthologous relationships.

Methodology:

Maps nucleotide and amino acid sequences using profile hidden Markov models (HMMs) and applies a best reciprocal hit (reciprocal search) strategy to assign sequences to the globally best matching clusters of orthologous genes.