SATCHMO-JS

SATCHMO-JS estimates protein multiple sequence alignments (MSAs) and phylogenetic trees using hidden Markov models to reconstruct protein evolutionary relationships.

Key Features:

• Simultaneous Estimation: Performs joint estimation of protein MSAs and phylogenetic trees.
• Input and Output Formats: Accepts protein sequences in FASTA format and produces an MSA and a phylogenetic tree as outputs.
• Algorithmic Approach: Extends the original SATCHMO algorithm with a divide-and-conquer strategy that jump-starts alignment at higher points in the phylogenetic tree to reduce progressive all-versus-all HMM scoring and alignment.
• Hidden Markov Models (HMMs): Employs HMMs for scoring and alignment during simultaneous MSA and tree estimation.
• Improved Accuracy: Demonstrated statistically significant improvements in alignment accuracy over MUSCLE, MAFFT, ClustalW, and the original SATCHMO using 983 structurally aligned pairs from the PREFAB benchmark dataset.

Scientific Applications:

• Phylogenomics: Reconstructs protein family phylogenies for phylogenomic analyses.
• Evolutionary Biology: Infers evolutionary relationships and patterns of protein evolution.
• Functional Inference: Supports inference of functional relationships among proteins via improved alignments and trees.
• Structural Analysis: Aids analysis of structural characteristics of protein families by providing accurate MSAs.
• Large-scale Genomic Studies: Applicable to large-scale genomic or proteomic datasets requiring precise alignment and tree construction.

Methodology:

Performs simultaneous estimation of MSAs and phylogenetic trees using hidden Markov models and a divide-and-conquer strategy that jump-starts alignment at higher nodes to reduce progressive all-versus-all HMM scoring and alignment.