COATi

COATi statistically aligns pairs of protein-coding sequences with codon-aware models that infer insertions and deletions both within and between codons.

Key Features:

• Codon-aware alignment: Accounts for indels within codons, addressing a limitation of traditional amino-acid- and codon-based aligners that assume indels occur only between codons.
• Complex insertion-deletion models: Supports statistical models for indel inference that capture biological variation in protein-coding sequences.
• Artifact handling: Processes sequencing artifacts and genome assembly errors including abiological frameshifts and incorrect early stop codons present in genomic data.
• Reduced data loss: Enables retention of more sequence data by generating accurate alignments that would otherwise be discarded.

Scientific Applications:

• Comparative genomics: Generates accurate alignments of orthologous protein-coding sequences for cross-species comparison, as demonstrated on human-gorilla orthologs.
• Phylogenetic inference: Provides improved sequence alignments as input for reconstructing evolutionary relationships.
• Ancestral sequence reconstruction: Supplies accurate aligned sequences needed for inferring ancestral protein and DNA sequences.
• Gene annotation: Supports accurate alignment-based identification and characterization of protein-coding genes.
• Functional genomics: Reduces erroneous conclusions caused by alignment errors from artifacts and incorrect indel modeling.

Methodology:

COATi uses statistical pairwise alignment with codon-aware scoring that explicitly models indels occurring both within and between codons.