Fig. 2

Haplotype network of global variation across ~ 10 kbs encompassing LTα and the 5′ region. Haplotype relationships were constructed using the median-joining algorithm implemented in Network 5.0 [38]. The resulting phylogeny consists of nodes (representing haplotypes), and links which connect the nodes in the shortest possible tree. The size of the nodes is proportional to the number of chromosomes with a given haplotype, and the colors within haplotypes (or nodes) indicate the frequency of that haplotype in pooled populations (such as, South Asians or East Asians). Distinct clusters of haplotypes are surrounded by circles/ovals with dashed lines, and the alleles that define the highest-frequency haplotype(s) are also given in each cluster. Underlined nucleotides indicate alleles that are inferred to be under selection and are shared with archaic hominins. Each haplotype cluster was arbitrarily labelled as Cluster I, Cluster II and Cluster III