Integrating Sequence- and Structure-Based Similarity Metrics for the Demarcation of Multiple Viral Taxonomic Levels

Viruses exhibit significantly greater diversity than cellular organisms, posing a complex challenge to their taxonomic classification. While primary sequences may diverge considerably, protein functional domains can maintain conserved 3D structures throughout evolution. Consequently, structural homology of viral proteins can reveal deep taxonomic relationships, overcoming limitations inherent in sequence-based methods. In this work, we introduce MPACT (Multimetric Pairwise Comparison Tool), an integrated tool that utilizes both sequence- and structure-based metrics. The program incorporates five metrics: sequence identity, similarity, maximum likelihood distance, TM-score, and 3Di-character similarity. MPACT generates heatmaps and distance trees to visualize viral relationships across multiple levels, enabling users to substantiate viral taxa demarcation. Taxa delineation can be achieved by specifying appropriate score cutoffs for each metric, facilitating the definition of viral groups, and storing their corresponding sequence data. By analyzing diverse viral datasets spanning various levels of divergence, we demonstrate MPACT's capability to reveal viral relationships, even among distantly related taxa. This tool provides a comprehensive approach to assist viral classification, exceeding the current methods by integrating multiple metrics and uncovering deeper evolutionary connections.