Geographic patterns of (genetic, morphologic, linguistic) variation: how barriers can be detected by using...

Area of Science:

• Population genetics
• Computational geometry
• Bioinformatics

Background:

• Traditional methods like spatial autocorrelation and regression analyze genetic-geographic distance associations but fail to pinpoint genetic barriers.
• Identifying abrupt changes in genetic variation is crucial for understanding population structure and evolutionary processes.

Purpose of the Study:

• To develop and implement a novel computational geometry approach for precise identification of genetic barriers.
• To enhance Monmonier's maximum difference algorithm with a significance test for robust barrier visualization.

Main Methods:

• Implementation of Monmonier's maximum difference algorithm in a new software package.
• Integration of a bootstrap matrices analysis for significance testing of identified barriers.
• Application of a multiple matrices approach to visualize variation patterns across different genetic markers.

Main Results:

• The software accurately identifies the location and direction of genetic barriers on geographic maps.
• Noise associated with genetic markers is visualized, highlighting areas with robust genetic barriers.
• The multiple matrices approach enables simultaneous visualization of variation patterns from different markers.

Conclusions:

• The improved Monmonier's method provides a reliable tool for detecting genetic barriers and understanding genetic landscapes.
• This approach is applicable to various genetic markers and can be extended to non-genetic distance matrix data.