Sets of complementary phenotypic relationships across five isolated populations: Traces of expression of different gene complexes?

Abstract

Sets of complementary phenotypic relationships across five isolated populations: Traces of expression of different gene complexes?The aim of this stud}, is to search for certain repeating phenolypic patterns, i.e. sets of complementary relationships across five isolated populations, which may represent the traces of expression of different genes or gene complexes. The study was conducted anaong isolates of five island populations of eastern Adriatic, Croatia, and the data were collected between 1979 and 1990. Selected phenotypic characteristics included measures of biological distances (e.g. anthropomelrical body and head distances, physiological, dermatoglyphic and radiogrammctric bone distances), while other examined traits included sociocultural (linguistic), bi0-cultural (migrational kinship) and genetic distances. The sample consisted of 6,286 examinees from 43 villages of five isolate populations. Correlations belween distance matrices based on examined traits were analyzed in each of five populations using Mantel's test of matrix correspondence, and factor analysis (rotated principal component) was then performed over obtained correlation matrices. The results showed that there were several consistent and significant correlations between some analyzed traits across all of the studied isolate populations, which might indicate their regulation by the shared gene complexes or genome regions. The analyses identified three main clusters of correlations in all five isolate populations: the first one containing anthropometric measures (body and head measures and physiological properties in both sexes), the second one containing geographic distance-related traits (migrational kinship, linguistic and genetic distances), and the third one containing dermatoglyphic properties and radiogrammetric bone measures in both sexes. The higher order varimax rotation over the matrix of factor correlations revealed that the primary source of variation within all five analyzed populations was not sex-related, but rather variable-specific.