Ancestry analysis using autosomal SNPs in northern South America, reveals interpretation differences between an AIM panel and an identification panel

•Ancestry evaluation using SNPforID 52plex is biased compared to subset 79 AIMs belonging Lace panel.•Ancestry evaluation generated with the identification panel must be interpreted with caution.•New SNPs panel with higher In for Native American populations would improvement ancestry assessment.•Convenience non-probability sampling, demographic previous study help to found Native American ancestry admixed populations.•Convenience non-probability sampling, historical, demographic tools in ancestry evaluation. Current human populations are studied to elucidate their ancestry composition and to obtain reference values for an array of genetic markers for forensic practice. This study compared the 79 ancestry informative markers (AIMs) panel with the SNPforID 52plex set used in forensic identification, using samples belonging to Continental Caribbean populations from Colombia with a high percentage of locals self-determined as Native American descendants. The results show a bias in the individual estimation made with the identification markers, which disregards the Native American ancestry component and overestimates the African ancestry component. Also, the analysis made with the Bayesian Classification Algorithm shows better likelihoods for individual assignment with AIMs than with SNPforID 52plex.