DCAlign

DCAlign aligns homologous biological sequences using co-evolutionary models and empirical priors to improve multiple sequence alignment accuracy by capturing conservation and co-evolution signals.

Key Features:

• Co-Evolution Models: Uses co-evolutionary models to detect and incorporate residue-residue evolutionary couplings across multiple sequence alignments.
• Informed Priors: Integrates empirical priors over variables modeling insertions and deletions to account for indel events in alignments.
• Computational Efficiency: Reduces computational demands through strategic pre-processing steps that streamline the alignment computation.
• Implementation in Julia: Implemented in the Julia programming language.

Scientific Applications:

• Protein Structure Prediction: Produces refined alignments that aid inference of structural motifs and residue-residue contacts for protein structure prediction.
• Phylogenetic Analysis: Provides refined alignments that better reflect evolutionary histories for phylogenetic tree reconstruction.
• Functional Annotation of Genes: Identifies conserved regions across species to support functional annotation of genes.

Methodology:

Integrates co-evolutionary information with empirical priors over insertion and deletion variables and employs pre-processing to reduce computational cost; implemented in Julia.