HaMStR: profile hidden markov model based search for orthologs in ESTs

Area of Science:

• Bioinformatics
• Computational Biology
• Evolutionary Biology

Background:

• Expressed Sequence Tag (EST) sequencing rapidly provides protein-coding sequences, crucial for organisms lacking complete genome data.
• ESTs are vital for evolutionary studies but lack efficient ortholog prediction methods, hindering their use in molecular systematics.
• Existing ortholog prediction methods struggle with EST data redundancy and often require a known species tree.

Purpose of the Study:

• To present a novel, automated approach for ortholog prediction in EST data.
• To overcome limitations of existing methods in handling EST redundancy and species tree dependency.
• To facilitate the use of ESTs in large-scale phylogenomic and evolutionary analyses.

Main Methods:

• HaMStR combines Profile Hidden Markov Model (HMM) searches with subsequent BLAST searches.
• The method extends existing ortholog clusters by incorporating sequences from additional taxa.
• It processes EST and protein sequence data for ortholog identification.

Main Results:

• HaMStR accurately identifies orthologs in EST data, yielding results consistent with methods using complete genomes.
• A case study on fungal phylogeny demonstrated HaMStR's utility in constructing informative datasets from ESTs.
• The approach effectively compiles orthologous gene sets from diverse protein-coding sequence data.

Conclusions:

• HaMStR provides a standardized method to augment predefined ortholog sets with EST data from new taxa.
• The approach is applicable to both EST and protein sequence data, offering a comprehensive tool for ortholog clustering.
• HaMStR facilitates various evolutionary studies, including molecular systematics and tracing the evolution of gene families.