The logic of human destiny was inevitable 1 million years ago

Robert Wright’s best book, Nonzero: The Logic of Human Destiny, was published nearly 20 years ago. At the time I was moderately skeptical of his thesis. It was too teleological for my tastes. And, it does pander to a bias in human psychology whereby we look to find meaning in the universe.

But this is 2017, and I have somewhat different views.

In the year 2000 I broadly accepted the thesis outlined a few years later in The Dawn of Human Culture. That our species, our humanity, evolved and emerged in rapid sequence, likely due to biological changes of a radical kind, ~50,000 years ago. This is the thesis of the “great leap forward” of behavioral modernity.

Today I have come closer to models proposed by Michael Tomasello in The Cultural Origins of Human Cognition and Terrence Deacon in The Symbolic Species: The Co-evolution of Language and the Brain. Rather than a punctuated event, an instance in geological time, humanity as we understand it was a gradual process, driven by general dynamics and evolutionary feedback loops.

The conceit at the heart of Robert J. Sawyer’s often overly preachy Neanderthal Parallax series, that if our own lineage went extinct but theirs did not they would have created a technological civilization, is I think in the main correct. It may not be entirely coincidental that the hyper-drive cultural flexibility of African modern humans evolved in African modern humans first. There may have been sufficient biological differences to enable this to be likely. But I believe that if African modern humans were removed from the picture Neanderthals would have “caught up” and been positioned to begin the trajectory we find ourselves in during the current Holocene inter-glacial.

Luke Jostins’ figure showing across board encephalization

The data indicate that all human lineages were subject to increased encephalization. That process trailed off ~200,000 years ago, but it illustrates the general evolutionary pressures, ratchets, or evolutionary “logic”, that applied to all of them. Overall there were some general trends in the hominin lineage that began to characterized us about a million years ago. We pushed into new territory. Our rate of cultural change seems to gradually increased across our whole range.

One of the major holy grails I see now and then in human evolutionary genetics is to find “the gene that made us human.” The scramble is definitely on now that more and more whole genome sequences from ancient hominins are coming online. But I don’t think there will be such gene ever found. There isn’t “a gene,” but a broad set of genes which were gradually selected upon in the process of making us human.

In the lingo, it wasn’t just a hard sweep from a de novo mutation. It was as much, or even more, soft sweeps from standing variation.

Mouse fidelity comes down to the genes

While birds tend to be at least nominally monogamous, this is not the case with mammals. This strikes some people as strange because humans seem to be monogamous, at least socially, and often we take ourselves to be typically mammalian. But of course we’re not. Like many primates we’re visual creatures, rather than relying in smell and hearing. Obviously we’re also bipedal, which is not typical for mammals. And, our sociality scales up to massive agglomerations of individuals.

How monogamous we are is up for debate. Desmond Morris, who is well known to many from his roles in television documentaries, has been a major promoter of the idea that humans are monogamous, with a focus on pair-bonds. In contrast, other researchers have highlighted our polygamous tendencies. In The Mating Mind Geoffrey Miller argues for polygamy, and suggests that pair-bonds in a pre-modern environment were often temporary, rather than lifetime (Miller is now writing a book on polyamory).

The fact that in many societies high status males seem to engage in polygamy, despite monogamy being more common, is one phenomenon which confounds attempts to quickly generalize about the disposition of our species. What is preferred may not always be what is practiced, and the external social adherence to norms may be quite violated in private.

Adducing behavior is simpler in many other organisms, because their range of behavior is more delimited. When it comes to studying mating patterns in mammals voles have long been of interest as a model. There are vole species which are monogamous, and others which are not. Comparing the diverged lineages could presumably give insight as to the evolutionary genetic pathways relevant to the differences.

But North American deer mice, Peromyscus, may turn to be an even better bet: there are two lineages which exhibit different mating patterns which are phylogenetically close enough to the point where they can interbreed. That is crucial, because it allows one to generate crosses and see how the characteristics distribute themselves across subsequent generations. Basically, it allows for genetic analysis.

And that’s what a new paper in Nature does, The genetic basis of parental care evolution in monogamous mice. In figure 3 you can see the distribution of behaviors in parental generations, F1 hybrids, and the F2, which is a cross of F1 individuals. The widespread distribution of F2 individuals is likely indicative of a polygenic architecture of the traits. Additionally, they found that some traits are correlated with each other in the F2 generation (probably due to pleiotropy, the same gene having multiple effects), while others were independent.

With the F2 generation they ran a genetic analysis which looked for associations between traits and regions of the genome. They found 12 quantitative trait loci (QTLs), basically zones of the genome associated with variation on one or more of the six traits. From this analysis they immediately realized there was sexual dimorphism in terms of the genetic architecture; the same locus might have a different effect in the opposite sex. This is evolutionarily interesting.

Because the QTLs are rather large in terms of physical genomic units the authors looked to see which were plausible candidates in terms of function. One of their hits was vasopressin, which should be familiar to many from vole work, as well as some human studies. Though the QTL work as well as their pup-switching experiment (which I did not describe) is persuasive, the fact that a gene you’d expect shows up as a candidate really makes it an open and shut case.

The extent of the variation explained by any given QTL seems modest. In the extended figures you can see it’s mostly in the 1 to 5 percent range. In Carl Zimmer’s excellent write up he ends:

But Dr. Bendesky cautioned that the vasopressin gene would probably turn out to be just one of many that influence oldfield mice. Though it is strongly linked to parental behavior, the vasopressin gene accounts for 6.7 percent of the variation in nest building among males, and only 2.9 percent among females.

The genetic landscape of human parenting will turn out to be even more rugged, Dr. Bendesky predicted.

“You cannot do a 23andMe test and find out if your partner is going to be a good father,” he said.

Sort of. The genetic architecture above is polygenic…but not incredibly diffuse. The proportion of variation explained by the largest effect allele is more than for height, and far more than for education. If human research follows up on this, I wouldn’t be surprised if you could develop a polygenic risk score.

But I don’t have a good intuition on how much variation in humans there really is for these sorts of traits that are heritable. I assume some. But I don’t know how much. And how much of the variance in behavior might be explained by human QTLs? Humans don’t lick or build nests, or retrieve pups. Also, as one knows from Genetics and Analysis of Quantitative Traits sexually dimorphic traits take a long time to evolve. These are two deer mice species. Within humans there may not have been enough time for this sort of heritable complexity of behavior to evolve.

There are a lot of philosophical issues here about translating to a human context.

Nevertheless, this research shows that ingenious animal models can powerfully elucidate the biological basis of behavior.

Citation: The genetic basis of parental care evolution in monogamous mice. Nature (2017) doi:10.1038/nature22074

Women hate going to India

For some reason women do not seem to migrate much into South Asia. In the late 2000s I, along with others, noticed a strange discrepancy in the Y and mtDNA lineages which trace one’s direct male and female lines: in South Asia the male lineages were likely to cluster with populations to the north an west, while the females lines did not. South Asia’s females lines in fact had a closer relationship to the mtDNA lineages of Southeast and East Asia, albeit distantly.

One solution which presented itself was to contend there was no paradox at all. That the Y chromosomal lineages found in South Asia were basal to those to the west and north. In particular, there were some papers suggesting that perhaps R1a1a originated in South Asia at the end of the last Pleistocene. Whole genome sequencing of Y chromosomes does not bear this out though. R1a1a went through rapid expansion recently, and ancient DNA has found it in Russia first. But in 2009 David Reich came out with Reconstructing Indian population history, which offered up somewhat of a possible solution.

What Reich and his coworkers found that South Asia seems to be characterized by the mixture of two very different types of populations. One set, ANI (Ancestral North Indian), are basically another western or northwestern Eurasian group. ASI (Ancestral South Indian), are indigenous, and exhibit distant affinities to the Andaman Islanders. The India-specific mtDNA then were from ASI, while the Y chromosomes with affinities to people to the north and west were from ANI. In other words, the ANI mixture into South Asia was probably through a mass migration of males.

But it’s not just Y and mtDNA in this case only. A minority of South Asians speak Austro-Asiatic languages. The most interesting of these populations are the Munda, who tend to occupy uplands in east-central India. Older books on India history often suggest that the Munda are the earliest aboriginals of the subcontinent, but that has to confront the fact that most Austro-Asiatic language are spoken in Southeast Asia. There was no true consensus where they were present first.

Genetics seems to have solved this question. The evidence is building up that Austro-Asiatic languages arrived with rice farmers from Southeast Asia. Though most of the ancestry of the Munda is of ANI-ASI mix, a small fraction is clearly East Asian. And interestingly, though they carry no East Asian mtDNA, they do carry East Asian Y. Again, gene flow mediated by males.

The same is true of India’s Bene Israel Jewish community.

A new preprint on biorxiv confirms that the Parsis are another instance of the same dynamic: The genetic legacy of Zoroastrianism in Iran and India: Insights into population structure, gene flow and selection:

Zoroastrianism is one of the oldest extant religions in the world, originating in Persia (present-day Iran) during the second millennium BCE. Historical records indicate that migrants from Persia brought Zoroastrianism to India, but there is debate over the timing of these migrations. Here we present novel genome-wide autosomal, Y-chromosome and mitochondrial data from Iranian and Indian Zoroastrians and neighbouring modern-day Indian and Iranian populations to conduct the first genome-wide genetic analysis in these groups. Using powerful haplotype-based techniques, we show that Zoroastrians in Iran and India show increased genetic homogeneity relative to other sampled groups in their respective countries, consistent with their current practices of endogamy. Despite this, we show that Indian Zoroastrians (Parsis) intermixed with local groups sometime after their arrival in India, dating this mixture to 690-1390 CE and providing strong evidence that the migrating group was largely comprised of Zoroastrian males. By exploiting the rich information in DNA from ancient human remains, we also highlight admixture in the ancestors of Iranian Zoroastrians dated to 570 BCE-746 CE, older than admixture seen in any other sampled Iranian group, consistent with a long-standing isolation of Zoroastrians from outside groups. Finally, we report genomic regions showing signatures of positive selection in present-day Zoroastrians that might correlate to the prevalence of particular diseases amongst these communities.

The paper uses lots of fancy ChromoPainter methodologies which look at the distributions of haplotypes across populations. But some of the primary results are obvious using much simpler methods.

1) About 2/3 of the ancestry of Indian Parsis derives from an Iranian population
2) About 1/3 of the ancestry of Indian Parsis derives from an Indian popuation
3) Almost all the Y chromosomes of Indian Parsis can be accounted for by Iranian ancestry
4) Almost all the mtDNA haplogroups of Indian Parsis can be accounted for by Indian ancestry
5) Iranian Zoroastrians are mostly endogamous
6) Genetic isolation has resulted in drift and selection on Zoroastrians

The fact that the ancestry proportion is clearly more than 50% Iranian for Parsis indicates that there was more than one generation of males who migrated. They did not contribute mtDNA, but they did contribute genome-wide to Iranian ancestry. There are wide intervals on the dating of this admixture event, but they are consonant oral history that was later written down by the Parsis.

So there you have it. Another example of a population formed from admixture because women hate going to India.

Citation: The genetic legacy of Zoroastrianism in Iran and India: Insights into population structure, gene flow and selection.
Saioa Lopez, Mark G Thomas, Lucy van Dorp, Naser Ansari-Pour, Sarah Stewart, Abigail L Jones, Erik Jelinek, Lounes Chikhi, Tudor Parfitt, Neil Bradman, Michael E Weale, Garrett Hellenthal
bioRxiv 128272; doi:

Genetic variation in human populations and individuals

I’m old enough to remember when we didn’t have a good sense of how many genes humans had. I vaguely recall numbers around 100,000 at first, which in hindsight seems rather like a round and large number. A guess. Then it went to 40,000 in the early 2000s and then further until it converged to some number just below 20,000.

But perhaps more fascinating is that we have a much better catalog of the variation across the whole human genome now. Often friends ask me questions of the form: “so DTC genomic company X has about 800,000 SNPs, is that enough to do much?” To answer such a question you need some basic numbers in your head, as well as what you want to “do.”

First, the human genome has about 3 billion base pairs (3 Gb). That’s a lot. But most of the genome famously doesn’t code for proteins. The exome, the proportion of the genome where bases directly translate into a protein accounts for 1% of the whole genome. That’s 30 million bases (30 Mb). But this small region of the genome is very important, as the vast majority of major disease mutations are found in the exome.

When it comes to a standard 800K SNP chip, which samples 800,000 positions across the 3 Gb genome, it is likely that the designers enriched the marker set for functional positions relevant to diseases. Not all marker positions are created equal. Though even outside of those functional positions there are often nearby SNPs that can “tag” them, so you can infer one from the state of the other.

But are 800,000 positions enough to make good ancestry inference? (to give one example) Yes. 800,000 is actually a substantial proportion of the polymorphism in any given genome. There have been some papers which improved on the numbers in 2015’s A global reference for human genetic variation, but it’s still a good comprehensive review to get an order-of-magnitude sense. The table below gives you a sense of individual variation:

Median autosomal variant sites per genome

When it comes to single nucleotide polymorphisms (SNPs), what SNP chips are getting at, an 800K array should get a substantial proportion of your genome-wide variation. More than enough for ancestry inference or forensics. The singleton column shows mutations specific to the individual.  When focusing on new mutations specific to an individual that might cause disease, singleton large deletions and nonsynonymous SNPs is really where I’d look.

But what about whole populations? The plot to the left shows the count of variants as a function of alternative allele frequency. When we say “SNP”, you really mean variants which exhibit polymorphism at a particular cut-off frequency for the minor allele (often 1%). It is clear that as the minor allele frequency increases in relation to the human reference genome the number of variants decreases.

From the paper:

The majority of variants in the data set are rare: ~64 million autosomal variants have a frequency <0.5%, ~12 million have a frequency between 0.5% and 5%, and only ~8 million have a frequency >5% (Extended Data Fig. 3a). Nevertheless, the majority of variants observed in a single genome are common: just 40,000 to 200,000 of the variants in a typical genome (1–4%) have a frequency <0.5% (Fig. 1c and Extended Data Fig. 3b). As such, we estimate that improved rare variant discovery by deep sequencing our entire sample would at least double the total number of variants in our sample but increase the number of variants in a typical genome by only ~20,000 to 60,000.

An 800K SNP chip will be biased toward the 8 million or so variants with a frequency of 5%. This number gives you a sense of the limited scope of variation in the human genome. 0.27% of the genome captures a lot of the polymorphism.

Citation: 1000 Genomes Project Consortium. “A global reference for human genetic variation.” Nature 526.7571 (2015): 68-74.

Why only one migrant per generation keeps divergence at bay

The best thing about population genetics is that because it’s a way of thinking and modeling the world it can be quite versatile. If Thinking Like An Economist is a way to analyze the world rationally, thinking like a population geneticist allows you to have the big picture on the past, present, and future, of life.

I have some personal knowledge of this as a transformative experience. My own background was in biochemistry before I became interested in population genetics as an outgrowth of my lifelong fascination with evolutionary biology. It’s not exactly useless knowing all the steps of the Krebs cycle, but it lacks in generality. In his autobiography I recall Isaac Asimov stating that one of the main benefits of his background as a biochemist was that he could rattle off the names on medicine bottles with fluency. Unless you are an active researcher in biochemistry your specialized research is quite abstruse. Population genetics tends to be more applicable to general phenomena.

In a post below I made a comment about how one migrant per generation or so is sufficient to prevent divergence between two populations. This is an old heuristic which goes back to Sewall Wright, and is encapsulated in the formalism to the left. Basically the divergence, as measured by Fst, is proportional to the inverse of 4 time the proportion of migrants times the total population + 1. The mN is equivalent to the number of migrants per generation (proportion times the total population). As the mN become very large, the Fst converges to zero.

The intuition is pretty simple. Image you have two populations which separate at a specific time. For example, sea level rise, so now you have a mainland and island population. Since before sea level rise the two populations were one random mating population their initial allele frequencies are the same at t = 0. But once they are separated random drift should begin to subject them to divergence, so that more and more of their genes exhibit differences in allele frequencies (ergo, Fst, the between population proportion of genetic variation, increases from 0).

Now add to this the parameter of migration. Why is one migrant per generation sufficient to keep divergence low? The two extreme scenarios are like so:

  1. Large populations change allele frequency very slowly due to drift, so only a small proportion of migration is needed to prevent them from diverging
  2. Small populations change allele frequency very fast due to drift, so a larger proportion of migration is needed to prevent them from drifting

Within a large population one migrant is a small proportion, but drift is occurring very slowly. Within a small population drift is occurring fast, but one migrant is a relatively large proportion of a small population.

Obviously this is a stylized fact with many details which need elaborating. Some conservation geneticists believe that the focus on one migrant is wrongheaded, and the number should be set closer to 10 migrants.

But it still gets at a major intuition: gene flow is extremely powerful and effective at reducing differences between groups. This is why most geneticists are skeptical of sympatric speciation. Though the focus above is on drift, the same intuition applies to selective divergence. Gene flow between populations work at cross-purposes with selection which drives two groups toward different equilibrium frequencies.

This is why it was surprising when results showed that Mesolithic hunter-gatherers and farmers in Europe were extremely genetically distinct in close proximity for on the order of 1,000 years. That being said, strong genetic differentiation persists between Pygmy peoples and their agriculturalist neighbors, despite a long history of living nearby each other (Pygmies do not have their own indigenous languages, but speak the tongue of their farmer neighbors). In the context of animals physical separation is often necessary for divergence, but for humans cultural differences can enforce surprisingly strong taboos. Culture is as strong a phenomenon as mountains or rivers….

The future shall, and should, be sequenced

Last fall I talked about a preprint, Human demographic history impacts genetic risk prediction across diverse populations. It’s now published in AJHG, with the same informative title, Human Demographic History Impacts Genetic Risk Prediction across Diverse Populations. Even though talked about this before, I thought it would be useful to highlight again.

To recap, GWAS is a pretty big deal, but only in the last 15 years or so. With genome-wide data researchers began to explore associations between diseases and population genetic variation. In some cases they discovered strong associations between characteristics and genetic variants, but in many casese it turned out that though a trait is highly heritable (e.g., schizophrenia) the causal variants are either not common or do not explain much of the variation in the poplation (or both).

But as the second decade of GWAS proceeds the sample sizes are getting larger, and researchers are moving from SNP-chips, with their various biases, to high quality whole-genome sequences. One of the major sorts of low hanging fruit in the minds of many people are rare variants. Basically SNP-chips are geared toward finding common variations within large populations, since they have a finite number of markers they are going to interrogate. Sequencing though is a comprehensive catalog of the genome in a relative sense. If you have high coverage (so you sample the site many times) you can easily discover rare mutations within an individual genome that makes them distinctive from almost the rest of the human race (these may be de novo mutations, or, they could be mutations private to their extended pedigree).

But context matters. Martin et al. find that confirmed GWAS hits in Europeans tend to exhibit decreased portability as a function of genetic distance. This isn’t entirely surprising, especially if rarer variants are part of the explanation. Rare variants usually emerged later in history, after the differentiation between geographic races.

A solution would be to have a diverse panel of populations in your studies. For many reasons this was not to be. Northwest Europeans are enormously enriched in current data sets. Martin et al. observe that recent this has diminished somewhat, from 95% European to less than 80%. But they observe that this is mostly due to the inclusion of “Asian” samples, as opposed to African and Native Americans, who remain as undererpresented as they did several years ago.

The African and Native American samples present somewhat different problems. The Native American groups are quite drifted due to bottlenecks. Likely they have their own variants due to the combined affects of mutation and selection through 15 to 20,000 years of isolation from other human populations. In contrast, the African groups have lots of diversity with a high time depth due to their ancestral histories, which are less subject to bottleneck effects. The prediction ability into Africans of current GWAS looks to be rather pathetic. This is reasonable because their diversity is poorly captured in Eurocentric study designs, and, they are more genetically diverged from Europeans than Asians are.

Ultimatley I think, and hope, this portability question will be of short term utility. As sequencing gets cheap, and studies become more numerous, we’ll fill in the gaps of understudied populations. Finally, ethics is above my paygrade, but I do hope those who demand a strenuous bar on consent keep in mind that that will result in slower growth of these study populations. Academics want to do a good job, but they also want to stay on the good side of IRB.

Citation: Martin, Alicia R., et al. “Human demographic history impacts genetic risk prediction across diverse populations.” bioRxiv (2016): 070797.

Why the future won’t be genetically homogeneous

While reading The Founders of Evolutionary Genetics I encountered a chapter where the late James F. Crow admitted that he had a new insight every time he reread R. A. Fisher’s The Genetical Theory of Natural Selection. This prompted me to put down The Founders of Evolutionary Genetics after finishing Crow’s chapter and pick up my copy of The Genetical Theory of Natural Selection. I’ve read it before, but this is as good a time as any to give it another crack.

Almost immediately Fisher aims at one of the major conundrums of 19th century theory of Darwinian evolution: how was variation maintained? The logic and conclusions strike you like a hammer. Charles Darwin and most of his contemporaries held to a blending model of inheritance, where offspring reflect a synthesis of their parental values. As it happens this aligns well with human intuition. Across their traits offspring are a synthesis of their parents. But blending presents a major problem for Darwin’s theory of adaptation via natural selection, because it erodes the variation which is the raw material upon which selection must act. It is a famously peculiar fact that the abstraction of the gene was formulated over 50 years before the concrete physical embodiment of the gene, DNA, was ascertained with any confidence. In the first chapter of The Genetical Theory R. A. Fisher suggests that the logical reality of persistent copious heritable variation all around us should have forced scholars to the inference that inheritance proceeded via particulate and discrete means, as these processes do not diminish variation indefinitely in the manner which is entailed by blending.

More formally the genetic variance decreases by a factor of 1/2 every generation in a blending model. This is easy enough to understand. But I wanted to illustrate it myself, so I slapped together a short simulation script. The specifications are as follows:

1) Fixed population size, in this case 100 individuals

2) 100 generations

3) All individuals have 2 offspring, and mating is random (no consideration of sex)

4) The offspring trait value is the mid-parent value of the parents, though I also including a “noise” parameter in some of the runs, so that the outcome is deviated somewhat in a random fashion from expected parental values

In terms of the data structure the ultimate outcome is a 100 ✕ 100 matrix, with rows corresponding to generations, and each cell an individual in that generation. The values in each cell span the range from 0 to 1. In the first generation I imagine the combining of two populations with totally different phenotypic values; 50 individuals coded 1 and 50 individuals coded 0. If a 1 and 1 mate, the produce only 1′s. Likewise with 0′s. On the other hand a 0 and a 1 produce a 0.5. And so forth. The mating is random in each generation.

The figure to the left illustrates the decay in the variance of the trait value over generation time in different models. The red line is the idealized decay: 1/2 decrease in variance per generation. The blue line is one simulation. It roughly follows the decay pattern, though it is deviated somewhat because it seems that there was some assortative mating randomly (presumably if I used many more individuals it would converge upon the analytic curve). Finally you see one line which follows the trajectory of a simulation with noise. Though this population follows the theoretical decay more closely initially, it converges upon a different equilibrium value, one where some variance remains. That’s because the noise parameter continues to inject this every generation. The relevant point is that most of the variation disappears < 5 generations, and it is basically gone by the 10th generation. To maintain variation in a blending inheritance model requires a great deal of mutation, the extent of which is just not plausible.

To get a different sense of what occurred in these two particular simulations, here are heat maps. The interval 0 and  1 now have shading in each sell. I am displaying only 50 generations here. The top panel is one without noise, while the bottom panel has the noise parameter.

The contrast with a Mendelian model is striking. Imagine that 0 and 1 are now coded by two homozygote genotypes, with heterozygotes exhibiting a value of 0.5. If all the variation is controlled by the genotypes, then you have three genotypes, and three trait values. If I change the scenario above to a Mendelian one than variance will initially decrease, but the equilibrium will be maintained at a much higher level, as 50% of the population will be heterozygotes (0.5), and 50% homozygotes of each variety (0 and 1). With the persistence of heritable variation natural selection can operate to change the allele frequencies over time without the worry that the trait values within a breeding population will converge upon each other too rapidly. This is true even in cases of polygenic traits. Height and I.Q. remain variant, because they are fundamentally heritable through discrete and digital processes.

All this is of course why the “blond gene” won’t disappear, redheads won’t go extinct, nor will humans converge upon a uniform olive shade in a panmictic future. A child is a genetic cross between parents, but only between 50% of each parent’s genetic makeup. And that is one reason they are not simply an “averaging” of parental trait values.

Mitochondrial Eve: a de facto deception?

The above image, and the one to the left, are screenshots from my father’s 23andMe profile. Interestingly, his mtDNA haplogroup is not particularly common among ethnic Bengalis, who are more than ~80% on a branch of M. This reality is clear in the map above which illustrates the Central Asian distribution my father’s mtDNA lineage. In contrast, his whole genome is predominantly South Asianform, as is evident in the estimate that 23andMe provided via their ancestry composition feature, which utilizes the broader genome. The key takeaway here is that the mtDNA is informative, but it should not be considered to be representative, or anything like the last word on one’s ancestry in this day and age.

As a matter of historical record mtDNA looms large in human population genetics and phylogeography for understandable reasons. Mitchondria produce more genetic material than is found in the nucleus, and so were the lowest hanging fruit in the pre-PCR era. Additionally, because mtDNA lineages do not recombine they are well suited to a coalescent framework, where an idealized inverted treelike phylogeny converges upon a common ancestor. Finally, mtDNA was presumed to be neutral, so reflective of demographic events unperturbed by adaptation, and characterized by a high mutation rate, yielding a great amount of variation with which to differentiate the branches of the human family tree.

Many of these assumptions are are now disputable. But that’s not the point of this post. In the age of dense 1 million marker SNP-chips why are we still focusing on the history of one particular genetic region? In a word: myth. Eve, the primal woman. The “mother of us all,” who even makes cameos in science fiction finales!

In 1987 a paper was published which found that Africans harbored the greatest proportion of mtDNA variation among human populations. Additionally, these lineages coalesced back to a common ancestor on the order of 150,000 years ago. Since mtDNA is present in humans, there was a human alive 150,000 years ago who carried this ancestral lineage, from which all modern lineages derive. Mitochondrial DNA is passed from mothers to their offspring, so this individual must have been a woman. In the press she was labeled Eve, for obvious reasons. The scientific publicity resulted in a rather strange popular reaction, culminating in a Newsweek cover where Adam and Eve are depicted as naked extras from Eddie Murphy’s Coming to America film.

The problem is that people routinely believe that mtDNA Eve was the only ancestress of all modern humans from the period in which she lived. Why they believe this is common sense, and requires no great consideration. The reality is that the story being told by science is the story of mtDNA, with inferences about the populations which serve as hosts for mtDNA being incidental. These inferences need to be made cautiously and with care. It is basic logic that a phylogeny will coalesce back to a common ancestor at some point. Genetic lineages over time go extinct, and so most mtDNA lineages from the time of Eve went extinct. There were many woman who were alive during the same time as Eve, who contributed at least as much, perhaps more, to the genetic character of modern humans today. All we can say definitively is that their mtDNA lineage is no longer present. As mtDNA is passed from mother to daughter (males obviously have mtDNA, but we are dead ends, and pass it to no one), all one needs for a woman’s mtDNA lineage to go extinct is for her to have only sons. Though she leaves no imprint on the mtDNA phylogeny, obviously her sons may contribute genes to future generations.

Prior to ancient DNA and the proliferation of dense SNP data sets scholars were a bit too ambitious about what they believed they could infer from mtDNA and Y lineages (e.g., The Real Eve: Modern Man’s Journey Out of Africa). We are in a different time now, inferences made about the past rest on more than one leg. But the legend of Eve of the mtDNA persists, not because of its compelling scientific nature, but because this is a case where science piggy-backs upon prior conceptual furniture. This yields storytelling power, but a story which is based on a thin basis of fact becomes just another tall tale.

All this is on my mind because one of the scientists involved with Britain’s DNA, Jim Wilson, has penned a response to Vincent Plagnol’s Exaggerations and errors in the promotion of genetic ancestry testing (see here for more on this controversy). Overall I don’t find Wilson’s rebuttal too persuasive. It is well written, but it has the air of sophistry and lawyerly precision. I have appreciated Wilson’s science before, so I am not casting aspersions at his professional competence. Rather, some of the more enthusiastic and uninformed spokespersons for his firm have placed him in a delicate and indefensible situation, and he is gamely attempting to salvage the best of a bad hand. Importantly, he does not reassure me in the least that his firm did not use Britain’s atrocious libel laws as a threat to mute forceful criticism of their business model on scientific grounds. A more general issue here is that Wilson is in a situation where he must not damage the prospects of his firm, all the while maintaining his integrity as a scientist. From what I have seen once science becomes a business one must abandon the pretense of being a scientist first and foremost, no matter how profitable that aura of objectivity may be. The nature of marketing is such that the necessary caution and qualification essential for science becomes a major liability in the processing of communicating. It’s about selling, not convincing.

Going back to Eve, Wilson marshals a very strange argument:

“The claim that Adam and Eve really existed, as you suggest, refers to the most recent common ancestors of the mtDNA and non-recombining part of the Y chromosome. I don’t agree that there is nothing special about these individuals: there must have been a reason why mitochondrial Eve was on the front cover of Time magazine in the late 80s!….

A minor quibble, but I suspect he means the Newsweek cover. More seriously, this line of argumentation is bizarre on scientific grounds. Rather, it is a tack which is more rational when aiming toward a general audience which might purchase a kit which they believe might tell them of their relationship to “Eve.”

In the wake of the discussion at Genomes Unzipped I participated in further exchanges with Graham Coop and Aylwyn Scally on Twitter, and decided to spend 20 minutes this afternoon asking people what they thought about mitochondrial Eve. By “people,” I mean individuals who are pursuing graduate educations in fields such as genetics and forensics. My cursory “field research” left me very alarmed. Naturally these were individuals who did not make elementary mistakes in regards to the concept, but there was great confusion. I can only wonder what’s going through the minds of the public.

Analogies, allusions, and equivalences are useful when they leverage categories and concepts which we are solidly rooted in, and transpose them upon a foreign cognitive landscape. By pointing to similarities of structure and relation one can understand more fully the novel ground which one is exploring. Saying that the president of India is analogous to the queen of England is an informative analogy. These are both positions where the individual is a largely ceremonial head of state. In contrast, the president of the United States and the queen of England are very different figures, because the American executive is not ceremonial at all. This is not a useful analogy, even though superficially it sees no lexical shift.

Who was Eve? A plain reading is that she is the ancestor of all humans, and more importantly, the singular ancestress of all humans back to the dawn of time. This is a concept which the public grasps intuitively. Who is mtDNA Eve? A woman who flourished 150,000 years ago, who happened to carry the mtDNA lineage which would drift to fixation in the ancestors of modern humans. I think this is a very different thing indeed. For purposes of poetry and marketing the utilization of the name Eve is justifiable. But on scientific grounds all it does is confuse, obfuscate, and mislead.

The fiasco that Vincent Plagnol stumbled upon is just a symptom of a broader problem. Scientists need to engage in massive conceptual clean up, as catchy phrases such as “mitochondrial Eve” and “Y Adam” permeated the culture over the past generation, and mislead many sincere and engaged seekers of truth. This is of the essence because personal genomics, and the scientific understanding of genealogy, are now moving out of the ghetto of hobbyists, enthusiasts, and researchers. Though I doubt this industry will be massive, it will be ubiquitous, and a seamless part of our information portfolio. If people still have ideas like mitochondrial Eve in their head it is likely to cloud their perception of the utility of the tools at hand, and their broader significance.

Buddy, can you spare some ascertainment?

The above map shows the population coverage for the Geno 2.0 SNP-chip, put out by the Genographic Project. Their paper outlining the utility and rationale by the chip is now out on arXiv. I saw this map last summer, when Spencer Wells hosted a webinar on the launch of Geno 2.0, and it was the aspect which really jumped out at me. The number of markers that they have on this chip is modest, only >100,000 on the autosome, with a few tens of thousands more on the X, Y, and mtDNA. In contrast, the Axiom® Genome-Wide Human Origins 1 Array Plate being used by Patterson et al. has ~600,000 SNPs. But as is clear by the map above Geno 2.0 is ascertained in many more populations that the other comparable chips (Human Origins 1 Array uses 12 populations). It’s obvious that if you are only catching variation on a few populations, all the extra million markers may not give you much bang for the buck (not to mention the biases that that may introduce in your population genetic and phylogenetic inferences).

To the left are the list of populations against which the Human Origins 1 Array was ascertained, and they look rather comprehensive to me. In contrast, for Geno 2.0 ‘ancestrally informative markers’ were ascertained on 450 populations. The ultimate question for me is this: is all the extra ascertainment on diverse and obscure groups worth it? On first inspection Geno 2.0′s number of SNPs looks modest as I stated, but in my experience when you quality control and merge different panels together you are often left with only a few hundred thousand SNPs in any case. 100-200,000 SNPs is also sufficient to elucidate relationships even in genetically homogeneous regions such as Europe in my experience (it’s more than enough for model-based clustering, and seems to be overkill for MDS or PCA). One issue that jumps out at me about the Affymetrix chip is that it is ascertained toward the antipodes. In contrast, Geno 2.0 takes into account the Eurasian heartland. I suspect, for example, that Geno 2.0 would be better for population or ancestry assignment for South Asians because it would have more informative markers for those populations.

Ultimately I can’t really say much more until I use both marker sets in different and similar contexts. Since Geno 2.0 consciously excludes many functional and medically relevant SNPs its utility is primarily in the domain of demographics and history. If the populations in question are well covered by the Human Origins 1 Array, I see no reason why one shouldn’t go with it. Not only does it have more information about biological function, but the number of markers are many fold greater. On the other hand, Geno 2.0 may be more useful on the “blank zones” of the Affy chip. Hopefully the Genographic Project results paper for Geno 2.0 will come out soon and I can pull down their data set and play with it.

Cite: arXiv:1212.4116

Unveiling the genealogical lattice

To understand nature in all its complexity we have to cut down the riotous variety down to size. For ease of comprehension we formalize with math, verbalize with analogies, and visualize with representations. These approximations of reality are not reality, but when we look through the glass darkly they give us filaments of essential insight. Dalton’s model of the atom is false in important details (e.g., fundamental particles turn out to be divisible into quarks), but it still has conceptual utility.

Likewise, the phylogenetic trees popularized by L. L. Cavalli-Sforza in The History and Geography of Human Genes are still useful in understanding the shape of the human demographic past. But it seems that the bifurcating model of the tree must now be strongly tinted by the shades of reticulation. In a stylized sense inter-specific phylogenies, which assume the approximate truth of the biological species concept (i.e., little gene flow across lineages), mislead us when we think of the phylogeny of species on the microevolutionary scale of population genetics. On an intra-specific scale gene flow is not just a nuisance parameter in the model, it is an essential phenomenon which must be accommodated into the framework.

This is on my mind because of the emergence of packages such as TreeMix and AdmixTools. Using software such as these on the numerous public data sets allows one to perceive the reality of admixture, and overlay lateral gene flow upon the tree as a natural expectation. But perhaps a deeper result is the character of the tree itself is torn asunder. The figure above is from a new paper, Efficient moment-based inference of admixture parameters and sources of gene flow, which debuts MixMapper. The authors bring a lot of mathematical heft to their exposition, and I can’t say I follow all of it (though some of the details are very similar to Pickrell et al.’s). But in short it seems that in comparison to TreeMix MixMapper allows for more powerful inference of a narrower set of populations, selected for exploring very specific questions. In contrast, TreeMix explores the whole landscape with minimal supervision. Having used the latter I can testify that that is true.

The big result from MixMapper is that it extends the result of Patterson et al., and confirms that modern Europeans seem to be an admixture between a “north Eurasian” population, and a vague “west Eurasian” population. Importantly, they find evidence of admixture in Sardinians, which implies that Patterson et al.’s original were not sensitive to admixture in putative reference populations (note that Patterson is a coauthor on this paper as well). The rub, as noted in the paper, is that it is difficult to estimate admixture when you don’t have “pure” ancestral reference populations. And yet here the takeaway for me is that we may need to rethink our whole conception of pure ancestral populations, and imagine a human phylogenetic tree as a series of lattices in eternal flux, with admixed nodes periodically expanding so as to generate the artifice of a diversifying tree. The closer we look, the more likely that it seems that most of the populations which have undergone demographic expansion in the past 10,000 years are also the products of admixture. Any story of the past 10,000 years, and likely the past 100,000 years, must give space at the center of the narrative arc lateral gene flow across populations.

Cite: arXiv:1212.2555 [q-bio.PE]