MAVID

MAVID performs global multiple sequence alignments of large genomic regions to identify conserved sequences and infer evolutionary relationships across species.

Key Features:

• High-Speed Performance: Aligns hundreds of kilobases in under a minute for large-scale genomic studies.
• Phylogenetic Analysis: Constructs phylogenetic trees from multiple alignments to provide insights into evolutionary relationships among sequences.
• Conserved Region Identification: Identifies conserved regions within alignments using the constructed phylogenetic trees as a basis.
• Progressive-Alignment Approach: Employs a progressive-alignment strategy incorporating maximum-likelihood inference for ancestral sequence prediction, automatic construction of guide-trees, protein-based anchoring to aid ab initio gene prediction, and constraints derived from a global homology map.
• Versatility and Robustness: Aligns highly divergent sequences as well as incomplete or unfinished sequences.
• Application Demonstrations: Demonstrated on a 1.8 Mb CFTR benchmark with 20 orthologous regions across marsupials, birds, fish, and mammals, on comprehensive alignments of all available HIV genomes, and on complete multiple alignments of entire genomes from humans, mice, and rats.

Scientific Applications:

• Evolutionary Biology: Reveals evolutionary relationships and divergence patterns among species through phylogenetic trees derived from alignments.
• Comparative Genomics: Detects conserved genomic elements across species to support cross-species comparisons and annotation transfer.
• Functional Genomics: Identifies conserved regions that may correspond to functional elements and supports ab initio gene prediction via protein-based anchoring.

Methodology:

Uses a progressive-alignment strategy with maximum-likelihood inference for ancestral sequence prediction, automatic guide-tree construction, protein-based anchoring, application of constraints from a global homology map, phylogenetic tree construction, and conserved-region identification.