Combining and congruence evaluation of phylogenetic signals from different genes based on geometric approach

• Published in: Vavilovskiĭ zhurnal genetiki i selekt͡s︡ii Vol. 20; no. 6; pp. 816 - 822
• Main Authors: Efimov, V. M., Kovaleva, V. Yu, Litvinov, Yu. N.
• Format: Journal Article
• Language: English
• Published: Siberian Branch of the Russian Academy of Sciences, Federal Research Center Institute of Cytology and Genetics, The Vavilov Society of Geneticists and Breeders 01.02.2017
• Subjects: dj-method; euclidean space; mtdna; nuclear dna; phylogenetics; sorex
• ISSN: 2500-0462; 2500-3259
• DOI: 10.18699/VJ16.153

A new Euclidean distance based algorithm is used for analysis of congruence and combining molecular genetic data. This approach is called geometric, since Euclidean distance satisfies all metric axioms and the points representing the sequences can be placed in a geometric space without distorting the mutual distances and can be endowed with the coordinates in this space. Geometricness of Euclidean distances allows to apply to molecular data methods of multivariate analysis, which are relevant for intra- and interspecies variability investigating, visualization of possible directions of evolution, combining data and evaluation of the congruence of phylogenetic signals. The algorithm is used for the analysis of more than 1500 nucleotide sequences of two nuclear (apoB, brca1) and two mitochondrial (co1, cytb) genes of 15 Palaearctic and Nearctic shrews species of genus Sorex (Soricidae, Eulipotyphla). All sequences of each gene are represented as a set of points in Euclidean space. Centroids of a set of points belonging to the same species are calculated. The matrix of Euclidean distances between the species centroids is calculated for each gene. Mantel test is applied to estimate pairwise similarity (congruence) of interspecies distances matrices relating to different genes. nDNA genes congruence is equal 0.961, mtDNA – 0.748. All matrices of the interspecies distances are combined into a joint matrix by weighing. Joint genetic space for all species is built by principal coordinate method from the joint matrix. Several variability directions reflecting evolutionary events of different scale are visualized in a joint genetic space. In addition, the joint matrix of interspecies distances is used for building a phylogenetic tree which is consistent with the zoological systematics accepted for today. This confirms the efficiency of our proposed method.