It's been a while since I've been in japan. As I've seen many threads concerning the origin of many east asian people on various asian forums, I am increasingly keen to delve into the division of east asian ethnicities essentially from scientific viewpoints. Consensus scientific conclusions seem to be that han chinese population who comprise more than 90% of chinese people have a high correspondece with the south east asian people such as Vietnamese, Cambodian, and others. Unfortunately, genetic studies on korean people are always incomplete due to the insufficient results, but many predict that korean population are related to the people in mongol, manchuria (north east china, where 10 million local population lives there), and has a relation to japanese due to the recent migrations to japanese archipelago. Japanese, on the other hand, are said to come from various places, south east asia, sibelia, china, korea, and polynesia, Oceania. In this thread, I would like to take a more scientific approaches in our pursuit of history. A key difference from other threads is a focus on genetics, and systematic methodology in our analysis of genetics. First, I will clarify our essential ingredients of the human population analysis by means of genetics. We have roughly three types of the approaches when examining the human population: (1) Microsatellite/RFLP in loci other than Y-chromosome/mtDNA. A rather too broad, and generally not useful. (2) Y-chromosome/mtDNA: Currently most credited study due to larger sample sizes. Non-recombining nature allows the easy analysis, and thus accessible to many genetics novices. *QTL are usually used. (1) is more prone to Linkage disequilibrium. (3) Complete genome approaches: Still premature at this stage. Number of population were never more than 300 in each group, thus difficult to make inferences. Other approaches which are in general more experimental nature, and has low credit from academia include HLA, Loci-homogeniety, random mating (HWP), simulation-based, etc. In conclusion, (2) is currently most useful. (3) would replace (2) when technological advance allows, but the results of (2) would still be viable. One important issue with (2) is that maternal lineage finding by mtDNA is not as reliable as with Y-chromosomes. mtDNA is arguably non-recombinant. Schwartz and Vissing 2002 showed that this would not be the case, and there are even recombinations between female and male mtDNA. However, this seemingly important finding was easily proven to unaffect the existing population studies. (1) The slight recombination factor we see is just too small to make impact to existing studies. (2) clonal assumption does not hold for the human population genetics and methodology But the male/female ancestry detection by means of mtDNA might be biological non-sense, because of the possible recombination leakage from female to male, and vice versa. Y chromosome has a characteristic of passing from father to orphan. It has 78 genes, coding for only 23 distinct proteins. mtDNA has 37 genes overall. Genetic studies make use of the non-recombining portion of the Y-chromosome. The mutation rate in mtDNA can even be 0.1% in one generation (as a result of shorter nucleotides), which means that in thousand years, we might get the huge biases, which we can usually not tolerate. Y is however only pointing male ancestral relationship, thus we need the mtDNA to find the support for the conclusion from Y to complement the missing story of female lineages. This is however questioned by the recombination events in mtDNA between male and female. However, the existing mtDNA results should be still a sound indicator comparing to the HLA, loci-homogeneity, RFLP/microsatellites, since the mutation rates in these loci are much higher, and thus even less credible. In this thread, I would like to shed some new light on the ethnicities of east asian people by citing the studies on Y-chromosomes, with a possible coupling of the results from mtDNA.