A New Theoretical Approach to Ancestry Estimation as Applied to Human Crania

Area of Science:

• Anthropology
• Human Variation Studies
• Biometrics

Background:

• Traditional anthropological methods for estimating human ancestry have remained largely unchanged since the 1960s.
• Current approaches often rely on predefined group labels, which may not accurately reflect the continuous nature of human genetic variation (clines).

Purpose of the Study:

• To propose and demonstrate a new theoretical framework for ancestry estimation that bypasses the need for group labels.
• To utilize unsupervised clustering methods to identify natural groupings within craniometric data.

Main Methods:

• A novel workflow was developed that allows data to structure into natural clusters, termed "morphogroups," without prior labeling.
• Hierarchical clustering was employed to determine the optimal number of clusters.
• Principal Components Analysis (PCA) was used to explore the identified morphogroups.
• Random Forest modeling was applied for multi-iteration classification based on morphogroups.

Main Results:

• The application of the proposed method to the Howells Craniometric Data Set identified three main morphogroups, with varying numbers of subclusters.
• Classification accuracy for morphogroups was consistently in the mid-90% range.
• Sex estimations achieved high accuracy, ranging from 93% to 100%.

Conclusions:

• The developed method offers a data-driven approach to ancestry estimation, moving beyond traditional, label-dependent classifications.
• This framework highlights the potential for new methodologies in understanding human variation and identity.
• The study encourages a re-evaluation of current practices in anthropological research regarding ancestry estimation.