However, both the A1 and A2 variants are found in sub-Saharan Africa, as the ALFRED data showed. So, the Taq1A SNP in itself could not be the N locus.
In addition, we looked at the distribution of the A1 and A2 variants in Eastern Europe:
This paper noted that the A1 variant was actually the ancestral one, and the paper's results did not show any particular association of A1 with presumed non-sanguine areas. If anything, the haplotype that includes A1 is referred to as of "East Asian" origin.
Nevertheless, the situation is complicated because of the possibility of "entanglement" of traits N and A (lack of trait A implies the presence of trait N, and vice versa). The Taq1A SNP could be part of a broader relevant haplotype.
One could make a case that the TaqIA is actually related to trait A, with the A2 variant coding for inhibition of aggression. Thus, homozygous Taq1(A1/A1) would be required for expression of trait A, and this would explain why 1) allele A1 is related to anti-social disorder, and 2) allele A2 is related to "calm".