List, I have for a longer time wondered how to be able to find autosomal mutations or autosomal haplotypes in easier ways than reading manually trough and tenthousand or more haplotypes to each other. I recently found out (or it seems) how to find your own unique haplotype or mutations/recent mutations with limited geographic or individual spread in the genome using software and methodogy originally made for finding genetic diseases. In this software scientists typically define a trait or the occurance of a disease in cases and compare these to the controls. If there is a very strong correlation between the diseases and the cases but not the controls you have detected a genetic disease. However nobody says you cant change "disease" with something else you want to investigate, so what happend if you define a individual as a "disease"? I phased a panel of 138 individuals with 41k SNP's in chromosome 1 resulting in 277 haplotypes where 1 individual was defined as having a disease trait for both haplotypes. The software quickly indicated trough haplotype clustering that 2 SNP allele was absolutely unique for this individual (or at least in this panel). Further 3 SNP was found to be found in in the "disease" individual and some others in various degree. The first SNP had very limited distribution within Scandinavia only found in 3 individuals where 2 haplotypes was found in the "disease" individual. The second SNP was found more widespread with 4 individuals found in both Europe and Scandinavia, also here matching with 2 haplotypes from the "disease" individual. The third SNP was found in 3 individuals found in Scandinavia and in Europe also here including 2 haplotypes from the "disease" individual. Then after having identified candidate SNP its time to reconstruct the actual haplotype and its ability to differentiate. Its cool to find unique SNP or unique haplotypes however they are very private or worthless in geneology if they dont show up in someone else. Lets therefor look at the first SNP that was also found among others than the "disease" individal, also here the software do the job:. SNP as 1 locus or haplotype scaled up from single locus to unique haplotype: C-> 120 of 277 haplotypes for 4 ind. CTC -> 58 of 277 haplotypes for 4 ind. TCCTC -> 3 of 277 haplotypes for 3 ind CCCTC -> 55 of 277 haplotypes for 1 ind. ATTCCTC -> 2 of 277 haplotypes for 2 ind. CTTCCTC -> 1 of 277 haplotypes for 1 ind. ACCCCTC ->55 of 277 haplotypes for 1 ind. GATTCCTC ->2 of 277 haplotypes for 2 ind. ACTTCCTC ->1 of 277 haplotypes for 1 ind. AACCCCTC ->1 of 277 haplotypes for 1 ind. And finally all haplotypes fully differensiated: GGCCAGATTCCTC -> 1 of 277 haplotypes for 1 ind. AGCCAGATTCCTC-> 1 of 277 haplotypes for 1 ind. AGCCAACTTCCTC -> 1 of 277 haplotypes for 1 ind. GGCTAAACCCCTC-> 1 of 277 haplotypes for 1 ind. The haplotype at its largest is only 74k or 0.074 Mb making these likely very stable haplotypes unlikely to recombine. So then only imagination and the genetic correlation with these sets of imigations sets the limit from here, a natural forward step would be as these kind of haplotypes are very stable to make autosomal haplotype trees to attempt tracing human migrations similar to Y-SNP and mitokondria SNP. Anders