DeSignate: detecting signature characters in gene sequence alignments for taxon diagnoses

摘要

Historically, taxonomic diagnoses are restricted to morphological characters distinguishing a particular taxon (the query group) from related taxa (the reference group). Best practice for taxonomic studies suggests an integrative approach combining morphological, molecular, ecological, and physiological data [ – ]. Previous suggestions for applying divergence cut-off values of gene sequences to discriminate and define taxa (threshold-based approach), however, are based on the overall dissimilarity and are not character-based, i.e., do not use distinct molecular characters for separation and characterization [ , ]. In the character-based approach, each position of an alignment represents a molecular character which may adopt different states in gene sequence data (e.g., nucleotides and deletions). Diagnostic molecular characters are included in taxon diagnoses (e.g., of protists [ ] or animals [ , – ]). However, data from (potentially) related taxa for comparison with the type species are often lacking or difficult to obtain [ ]. Furthermore, available data is frequently not added consistently to formal diagnoses [ , ], due to problems in, for instance, the definition of diagnostic molecular characters and the designation of their positions, as well as the lack of suitable tools. For a standardized designation of the position of diagnostic molecular characters in taxon diagnoses, a reference sequence alignment and/or a reference sequence are recommended, facilitating comparability and reproducibility [ ].