Exploring the relationships between genetic, linguistic and geographic distances in Bantu-speaking populations

Objectives The predominance of Bantu languages in sub-Saharan Africa has sparked a large debate over the processes through which they came to disperse over time and space-the "Bantu expansion." The overall genetic similarity shown by Bantu-speaking populations indicates that movement of people occurred too, but the extent of the correlation between genetics, linguistics and geography has been a matter of debate among scholars of different disciplines. In this work, we aim to investigate how genetic, linguistic and geographic distances relate to each other in Bantu-speaking populations. Methods We analyzed genome-wide SNP array data from a set of 37 Bantu and non-Bantu-speaking populations together with related linguistic and geographic data. Due to the complex demographic relationships resulting from events of admixture in the history of these populations, we develop and implement a method for controlling the signatures of admixture. Results Genetic distances were only minimally correlated with linguistic and geographic distances, possibly as the result of gene flow from neighboring groups into Bantu-speaking populations. When signatures of admixture are controlled for, the correlation of genetic data with linguistic and geographic distances significantly increases. Discussion The increase of the correlation between linguistic and genetic distances after the signatures of admixture are taken into account is in agreement with a scenario of spatial co-dispersal of languages and people. Additional specific cultural and demographic dynamics have probably further affected the relationship between language and genetics, which will be necessary to take into account when integrating multidisciplinary data to reconstruct the history of populations.