Genome, diversity, and origins: the Y chromosome as a storyteller.

A of human genome variation may focus on one of two possible goals: understanding the genome region under study or solving historical and evolutionary questions specific to the population(s) analyzed. Understanding of variation of a given genome region has a genetic interest because it is a consequence of the dynamics of the genome and thus the evolutionary forces (mutation, selection in its varieties, drift, recombination, . . . ) may be understood. It is thus a way to understand the mechanisms that produce variation in the genome. On the other hand, when genetic variation is being analyzed, random individuals from specific populations offer the possibility to trace back their origin beyond the limitation of the genome region sampled. The goal then may be the evolutionary reconstitution of specific populations or, for a global sampling, the origin of our species. The underlying idea is very simple: if we are able to trace back the coalescence of genomic regions from an ample worldwide sample, we can infer the phylogeny of humans. The rationale can be sketched as follows: (i) The evolutionary dynamics of the genome region is known in pattern and tempo. (ii) From the extant variation the past can be inferred and the coalescence process reconstructed; in some cases just part of the genetic information is being used, and analyses do not fully exploit the information contained in the data. (iii) Translate the genetic process into a process of individuals that reproduce (the population), and the time and place of the ancestral individual (carrying the ancestral genome) may be recognized. (iv) If the geographic distribution of derivative genetic stages is known, the expansion process may be dissected as migratory waves and events. (v) The structure of the genetic variation may reveal demographic characteristics such as population size and subdivision as well as ancient dynamics, such as expansions. This simple pathway is not easy. Inferences from molecules to populations are not straightforward, and there have been recurrent worries on what was being analyzed, either the genes or genomic regions on one hand or the individuals, populations, or species on the other. There have been worries concerning the accuracy of our knowledge of genome dynamics, worries concerning the ability and power to detect specific processes and disentangle cases where more than one mechanism may have produced similar genetic patterns, and worries concerning the appropriateness of evolutionary models needed for the inference. And finally there have been worries from anthropologists who do not perceive the interface between the evolutionary biology of a species and that of tiny fragments of DNA, usually in noncoding regions, worries surrounding a fast-developing field, heir to classical population genetics, with brilliant novelties but also eager to get headlines. In this context, two related papers in this issue (1, 2) analyze an ample set of sequences of the Y chromosome to address the issue of human origins. Their main objective is to propose a novelty: the common origin of present humans is not as old as previously believed, but should be much more recent, around 50,000 years ago (ya), the age of coalescence of the Y chromosomes surveyed (2) and in congruence with an onset of expansion of around 30,000 ya (1). Moreover, both studies corroborate the existence of a substantial (or exponential) population growth and an African origin for modern humans, in accord with most other genetic data. The newly proposed age is much younger than dates consistent with nuclear and mitochondrial DNA (mtDNA) (see references in ref. 2), where the minimum coalescent age was obtained for mtDNA at around 150,000 years and much older for nuclear genes, except for the present case. Similar results have been obtained through other genetic approaches, such as short tandem repeat (STR, also called microsatellite) variation (3), where a figure of 156,000 years for the deepest split in human phylogeny was obtained, or Alu insertions (4), with a proposed age of 137,000 years for the time of separation of African versus non-African populations. What seems more surprising is the discrepancy with the results of nine diallelic polymorphic sites on the Y chromosome (5), where the analysis with the same methods as those in ref. 2 gave a figure of around 150,000 ya for the coalescence of the variation. The new analysis (2) on the same chromosome gives only one-third of the time. It is thus an interesting new proposal not only in human evolution but also in human evolutionary genetics. It could mean that a population with modern characteristics had to exist in Africa 50,000 ya and spread later to Eurasia and the rest of the world. Besides the concordance of this proposal with archaeological evidence, there are more strictly genetic issues to be discussed, namely the limitations of the theory and of the genetic data, the interpretation of the variation, and possible specific properties unique to the Y chromosome. New ideas coming from genetics in human evolution have had very different fates, and some cases are remembered for having proposed a paradigm shift strongly attacked by paleoanthropologists but later shown to be correct. This was the case for the time depth of the hominid branch, as a separate group from our close relatives, the chimpanzee and the bonobo. In most cases the process either to acceptance or rejection goes with a lively debate in which scholars from very different disciplines enter the fray with non-mutually intelligible languages. A proposal like the present one (1, 2), stemming from genetics, has implications in several fields, from which specific clarifications may be asked: (i) Evolutionary genetics: Are conclusions about population fully supported or may there be a bias due to the genomic region under study? Does all of the genome tell the same story? (ii) Mathematical genetics: May the models used be safely applied to the real world, apart from their theoretical elegance? What are the implications of violation of the theoretical assumptions? (iii) Human paleontology: Is the evolutionary history of humans as interpreted from the hard evidence (the fossils) compatible with the new genetic proposals? (iv) Archaeology and Paleodemography: Since cultural innovations are at the base of population expansions, are time and mode of cultural change correlated