SPA

Spatial Ancestry analysis (SPA) is a method for predicting ancestry or where an individual is from using the individual's DNA. Accurately modeling ancestry is an important step in identifying genetic variation involved in disease.

SPA was created by Wen-Yun Yang, John Novembre, Eleazar Eskin and Eran Halperin

For general questions, comments or suggestions about SPA, please email Eleazar Eskin (eeskin (AT) cs.ucla.edu).

Convergence

SPA follows an alternative optimization strategy to maximize the likelihood of genotype date. The alternation is between computing the position of individuals and computing a slope function to capture allele frequency change over space. The below videos show how SPA converge to true geography of Europe and world globe from random initialization of individual positions.