The quantitative genetics of epidermal ridges using multivariate maximum likelihood estimation

Finger prints have been used for identification purposes for many hundreds of years. They were first used as a secure signature for financial agreements by merchants and rulers, later employed for forensic identification by law enforcement officials, and more recently adopted by anthropologists for estimating population relationships. Anthropologists believed that the dermatoglyphic observations accurately reflected the genetic relationships among observed populations due to their early development in utero, high heritability, and selective neutrality. A highly heritable and neutral trait would reflect the genetic relationships between populations more accurately. These qualities have prompted several anthropologists to suggest that dermatoglyphic traits reflect population relationships more accurately than anthropometrics or single-locus markers. While the quantitative genetics of anthropometrics and implications for estimated population relationships have been ascertained recently, these have not been investigated adequately for dermatoglyphics.

Dermatoglyphic traits appear to have a significant genetic component, though many aspects are uncertain. Holt firmly established the idea that dermatoglyphic traits, especially TRC (Total Ridge Count, the sum of the greater of the radial and ulnar ridge counts of each finger), were highly heritable with little environmental or nonaddictive variation. Holt interpreted population differences in dermatoglyphics as evidence for natural selection, but many assumed selective neutrality. More recent results suggest that Holt's models were inadequate and that the genetic qualities of finger ridges include major gene, dominance, epistasis, and maternal effects. Like Holt, later researchers frequently analyzed composite variables, such as TRC (Total Ridge Count). TRC was used because of its more normal distribution and computational ease.

The inheritance of TRC alone does not explain the genetic properties of dermatoglyphics adequately, however. As Weninger and others pointed out, TRC simplified what could be vastly different finger configurations. At the level of the individual fingers, results from factor analysis of phenotypic values strongly suggested a "field effect" during the embryonic formation of epidermal ridges. Additionally, there has been little research concerned with the genetics of toe ridge counts, so a more comprehensive view of epidermal ridge formation in humans is still lacking.

Previous quantitative genetic analyses of dermatoglyphics have been limited to relatively small data sets and pedigrees that were divided into parent-child regressions and sibling-sibling correlations. To date, the largest dermatoglyphic sample with pedigree data was collected from over 800 German and Austrian families by Heinz Brehme and was analyzed here for the first time. The data comprise samples from normal families, families with dizygotic twins, and families with monozygotic twins. All of these family configurations were analyzed using the method of maximum likelihood estimation.

The results of this study strongly confirm the additive genetic basis of dermatoglyphic variability as well as the role of dominance and a "field effect" in the development of epidermal ridges in the fingers and toes. The contribution of additive genes to finger ridge counts was found to be much lower than previously estimated. This result was due to larger samples, better methodologies, and the avoidance of summary variables such as TRC, which were shown empirically to inflate the additive component. Additionally, finger ridge counts, palm ridge counts, and toe ridge counts are largely under separate genetic control.

The components of dermatoglyphic variation are best understood developmentally. Ridge count heritabilities are lower due to significant dominance and environmental effects, but also to a high level of developmental integration, as shown by high genetic correlations among variables. Epidermal Growth Factor, with its widespread action on the developing fetus and the epidermis, may be an especially important regulatory hormone for developmental integration.

Differences between the additive genetic and phenotypic variance-covariance matrices produced disparities in population relationships using phenotypic and additive genetic values derived from the data with pedigrees. Dermatoglyphic phenotypes do not accurately reflect the genetic relationships among these populations. When analyzing population structure, dermatoglyphic traits cannot be used as a surrogate for genetic data, and will produce distorted relationships among populations