(a) Small coloured circles with a matching colour to geographical regions represent the 54 reference points used for GPS predictions. Each circle represents a geographical point with longitude and latitude and a certain admixture proportion. The insets provide magnification for dense regions. (b) GPS individual assignment based on 54 points. Individual label and colour match their known region/state/country of origin using the following legend: BE (Bermudian), BU (Bulgarian), CHB (Chinese), DA (Danish), EG (Egyptian), FIN (Finnish), GO (Georgian), GR (German), GK (Greek), I-S/N/W/E (India, Southern/Northern/Western/Eastern), IR (Iranian), ID/TSI (Italy: Sardinian/Tuscan), JPT (Japanese), LWK (Kenya: Luhya), KU (Kuwaiti), LE (Lebanese), M-O/B/N/D/T (Madagascar: Antananarivo/Ambilobe/Manakara/Andilambe/Toliara), X-G/H/M (Mexico: Guanajuato/Hidalgo/Morelos), MG (Mongolian), N-S/K/H/T (Namibia: Southeastern/Kaokoveld/Hereroland/Tsumkwe), YRI (Yoruba from West African), P-C/N (Papuan: Papua New Guinea/Bougainville-Nasioi), PH/PEL (Peruvian: Highland/Lima), PR (Puerto Rican), RO (Romanian), CA (Northern Caucasian), R-M/T/A (Russians: Moscow/Tatarÿ/Altaian), S-J/U/S/K/ (RSA: Johannesburg/Underberg/Northern Cape/Free State), IBS (Iberian from Spain & Portugal), PT (Pamiri from Tajikistan), TU (Tunisian), UK (British from United Kingdom), VA (Vanuatu), KHV (Vietnam). Note: occasionally all samples of certain populations (for example, Vietnamese) were predicted to the same spot and thus appear as a single sample.