How to split P3 samples into two subgroup when estimating the admixture proportion

[vrmyang]

Good morning,

My genome wide allele frequencies result are shown below, and my questions is how do I split the P3 sample into two subgroups ?

scaffold position pop1 pop2 pop3 pop0
Chr01 9440 0.0 0.0 0.0156 0.0
Chr01 9454 0.0 0.0 0.0156 0.0
Chr01 9465 0.0333 0.0 0.0 0.0
Chr01 9495 0.0385 0.0 0.0 0.0
Chr01 9496 0.0385 0.0 0.0 0.0
Chr01 9513 0.2333 0.0729 0.1176 0.0
Chr01 9514 0.0 0.0 0.0303 0.0
Chr01 9530 0.1 0.0 0.0147 0.0
Chr01 9549 0.0667 0.0 0.0 0.0

Could you provide any clue about how to solve this problem ?

Best wishes,
vrmyang