Haplotypes are particular combinations of genetic variants in the same chromosome that travel together and remain relatively stable through various generations. These DNA arrangements have become central actants (Latour 2004) in the search for medically significant SNP’sFootnote 22 (single nucleotide polymorphisms); as well as powerful entities involved in the delimitation of ancestry, population’s temporality and variability.
As medical and cost/benefit devices, haplotypes, are pivotal to find meaningful genetic variations related to disease, without having to read the whole genome of a patient. For molecular anthropology, genetic diversity (i.e. haplogroups, haplotypes, mutation rates, AIM’s and SNP’s)Footnote 23 has been linked to populations’ temporality, ethnic boundaries, ancestral lineages (like in the Cohanim haplotype; Abu El-Haj 2004; Thomas et al. 1998) and evolutionary narratives (Sommers 2008).
To understand the role of haplotypes in contemporary genomics, we need to understand first the notion of Linkage Disequilibrium (LD), understood as the probability that a combination of certain genetic polymorphisms is non-randomly associated at two or more loci (Slatkin 2008); in population genomics patterns of LD are strongly related to a population history. Events such as genetic drifts, and bottlenecks transform the patterns of linkage disequilibrium; a drastic reduction in population size (bottlenecks), generally make some haplotypes disappear, increasing LD as a result.
Therefore haplotype length, and haplotypic diversity, are linked to population’s temporality “Haplotype blocks vary somewhat among human populations—they tend to be somewhat shorter in African populations (Slatkin 2008; HapMap 2003, 2009)”.Footnote 24 Haplotype blocks with less variabilityFootnote 25 and higher LD, are a sign of a younger population, while haplotypes with lower LD, and higher variability are read as a sign of an older population; since recombination necessarily implies more time. Therefore in molecular anthropology, Africans are the ones understood as the oldest population, using as a starting point anthropological and paleontological studies (also see: Sommers 2008).
For many of our informants haplotypes are incontrovertible proofs of continental/ethnic origin, which cannot be misleading or ambivalent. Yet haplotypes and their interpretation remain confusing and distant for many “experts” in the field of genomic medicine, who are not interacting with them on a daily basis (i.e., in the laboratory). Even for some scientists and physicians working inside the INMEGEN, the interpretation and constitution of haplotypes, remains obscure. Dr. Max, a physician with some training in medical genomics-in charge of top administrative functions inside the INMEGEN—told me that: “haplotypes are constituted of little coloured squares, which represent different continental groups... like the red ones are Indigenous, the grey ones are European... and so on.... you know. The computer gives you that information right away!!”
As a matter of fact those “red, grey and white ethnic squares” represent linkage disequilibrium (LD or a non random association of two or more alleles, in this case SNP’s). Of course these squares are not informative of ethnicity or biogenetic identities by themselves. As you can notice in the graphic representation of haplotype blocks, the colour of the little squares, change according to the higher or lower probability of finding any combination of two SNP’s together (e.g. red = high probability; white/grey = low or very low probability). The patterns of linkage disequilibrium (LD) differ from population to population,Footnote 26 providing a tool for researchers interested in identifying specific markers (known as tag SNP’s) that may act as reporters of other highly associated polymorphisms (Frazer et al. 2009).
As an example in Fig. 1b, the middle haplotype (block 2), covering SNP’s 10 to 22, depicts a high (LD) between SNP’s 10, 11, 12, so by genotyping any of these three SNP’s you will immediately know the other two. The first two combinations are CTA, and the combination of the other two haplotypes is TCG. So if you genotype SNP 12, you will know that depending on the allele, either A or G, it will be accompanied by CT or TC respectively. Patterns of LD (the basis for the construction of haplotypes) are also used as arguments in favour of mapping certain populations, since their genetic characteristics—as larger chromosomal areas—could make the search for medically relevant SNP’s easier or even feasible (Alkes et al. 2007; Silva-Zolezzi et al. 2009):
Recently admixed populations such as African-Americans and Latino ethnic groups are known to have areas of LD (linkage disequilibrium) that can extend over large chromosomal regions due to allele frequency differences between the ancestral populations. This increased LD among admixed populations can facilitate mapping complex traits in an approach generally referred to as admixture mapping (cf. Burchard in Fullwiley 2008: 716).
The previous quote used to endorse admixture mapping, is not mentioning that compared to Europeans, one of the “ancestral” ethnic groups of African-Americans, the admixed population has more haplotypic diversity, and shorter haplotype blocks (see: Gabriel et al. 2002: 2226–28). Therefore the areas of LD of African-Americans are a lot more similar to Yoruba than to European samples. If we extend this comparison to other ancestral populations, such as the Asian (composed of Japanese & Chinese samples of the HapMap), we would find again that the admixed Afro-Americans have shorter haplotype blocks. Consequently in the case of Afro-Americans, the idea that their LD can extend to large chromosomal areas, will be completely misleading except when compared to Yorubas, the most diverse population of the HapMap.Footnote 27
We should be careful not to naturalise time and admixture too fast, without being aware of its implications, or that certain assertions go against most of the specialised literature on the topic (Zhu et al. 2004; Gabriel et al. 2002; Wall and Pritchard 2003). Since recombination is not only a temporal phenomenon in relation to the historic-sexual mating of the “ancestral populations”, but also a question of genetic/hereditary time, the notion of “Mextizo” haplotypes in the next section will better illustrate this temporal duality.
The construction and delimitation of haplotypes is mediated through algorithms, probability and statistical tests (Gabriel et al. 2002). The contested boundaries of haplotype blocks (Zhu et al. 2004), and the questioning of their fundamental assumptions (Terwilliger and Hiekkalinna 2006) is one of many examples showing the controversial and fluctuating properties of human variation after the HGP. The production of genethnicities in human genomics is not fixed, it dwells on relative and probabilistic frequencies of certain genotypes; yet their circulation and translation as mediators between bio-genetic and ethno-racial identities in Mexico draws heavily from the popular notion of a dominant Mestizo identity. Mestizaje functions as a cultural filter to make genomic science meaningful in many ways (i.e. to support the project, to claim sovereignty (Schwartz 2008, 2009) or to question its representativeness).
Indigenous and “Mextizo” haplotypes
Inside INMEGEN Population Genomics laboratory, the existence of a “Genomic Map of Mexicans” fades away when it comes to talk to talk about the “Diversity Project”. Whenever the first name is used, it is to make reference to its political or marketing connotations. In comparison to the unitary Mestizo narrative, the way in which diversity is represented in the strong visual and probabilistic laboratory culture, comprises multiple concomitant ways to visualise admixture. Mestizos can become a cline along two clusters; one representing Europeans (CEU: coming from 30 US trios with central and western European ancestries) and the other cluster representing indigenous or Amerindians (AMI: coming from 30 Zapotecos). Despite the fact that studying indigenous genetic contribution (traduced in the lab as AMI-Amerindian ancestry-) is fundamental to the emergence of the Genomic Mestizo, it can only be understood in comparison to other “ancestral” populations (i.e. Asian (EA), African (YRI) and European(CEU) as represented in the international HapMap).
Mestizos can also be presented as a specific composite of ancestral percentages “in this model their mean ancestries (SD±) were 0.552 ± 0.154 AMI, 0.418 ± 0.155 EUR, 0,018 ± 0.035 for AFR and 0.012 ± 0.018 EA (Silva Zolezzi et al. 2009:2)”. The Mestizo category can also be disaggregated to compare differential contributions of ancestry in different states of the country, differences in heterozigocity and haplotype diversity; or it can be put together to show common variability and AIM’s (ancestry informative markers). Most importantly the majority of these molecular movements fit into existing knowledge, historical records (African Slaves entry points) and known indigenous population densities (Silva-Zolezzi et al. 2009:5).
In the field of genomics, genethnic distinctions are differentials in the frequencies of certain genetic variants between studied (i.e., mapped) populations (Fig. 1c, presents the principal vectors of variation). One curious (even anecdotal) piece of information in the laboratory related to haplotypic distinctions between indigenous and Mestizos, can reveal the tactics to articulate natural and social orders in Mexican Genomics arena. Irma Silva-Zolezzi explains the indigenous contribution and its relation to diversity, as follows: “It is kind of paradoxical that by the presence of indigenous ancestry, the variability of Mestizo populations is reduced in genomic terms, when compared to European populations”.
The lesser haplotypic diversity coming from indigenous genetic heritage confers some of its specific genetic characteristics to the Mestizo population; like unique allelic frequencies, or the 89 private SNP’s not found in any other population of the International Hap Map (Silva-Zolezzi et al. 2009:4). One of the curious characteristic of “Mestizo” haplotypes is their counter intuitive bio-temporality. Even though they are historically the youngest population in the country (Mestizo genealogy can be roughly traced to 500 years ago, with the symbolic and reproductive alliance between la Malinche and Hernan Cortes), their haplotypes are more diverse and (LD) areas are shorter than indigenous ones.
We had always considered this discordance interesting. Yet it was not until a discussion in the laboratory on the 15th of May, 3 days after the presidential presentation of “The Map of the Mexicans’ Genome”Footnote 28; that we really acknowledged the importance of such discordance to understand the re-articulation of population genomics with the Mestizo identity.
After the Friday’s seminar in the laboratory, a very interesting discussion on variability and temporality began. The authors of this paper were discussing how 90 Yorubas could be representative of the most diverse population in the world—Irma Silva-Zolezzi a little bit mad—answered to the question of her co-author: “so now you are going to play Dr. Elias game around sampling and representativeness right?”Footnote 29 We laughed, and she continued to explain to all of us the difficulties of making a haplotype map of African populations:
Making a traditional map of African diversity would be incredibly difficult, and might not even cover a good deal of the African diversity...Yet Africans—she emphasized—might become interesting as the new target of commercial/medical genomics, since populations with larger haplotypes have helped researchers to identify genomic regions in which to re-sequence genotypes and refine specific candidate SNP’s; and the smaller haplotypes of Africans could make the search easier.
Leonardo, the bio-informatician interrupted the chat, saying “yes the boxes in the African genome are smaller, and as populations become younger the haplotypes (boxes) are bigger”. Irma and I recovered the example of mestizo populations in which the haplotypes are smaller than the indigenous ones, going against the common sense of the mestizo as a younger population. It was precisely in the middle of the conversation that I remembered this piece of the interview; I had with Dr. X (pseudonym) a top Scientist/Politician of the Mexican genomic community almost 10 months before the discussion in Laboratory:
Dr. X—They are, they are... pure indigenous, let’s say... how do I know? We already ran their genome, and we could take those that are (pause) let me say... pure... in the sense that...that... they are not mixed. Because you can identify those that are not mixed from the ones that are mixed...
ES—With the Haplotypes...?
Dr. X—Of course!!...you say here no, no... The ones that are already mixed, since they are recent admixtures, then the blocks are very big, when the genome is mixed (clapping) you find yourself with very big blocks. If many generations have passed and the genome has been recombined, the blocks are very little, so they are very difficult to identify. But 500 years, identifying blocks... but them, if they are mixed, they mixed two or three generations ago, so the admixture is very grotesque (clapping)!!! Imagine that the Indigenous were the brown genome, and the European, Mestizos or whoever is green genome and you mix as they are mixed 50% and 50% (clapping) it is very obvious!!! You notice they are... and if they mixed four generations ago, it is the same, the difference is too grotesque!! But if you go with the Africans that have so many generations the blocks are so little that you cannot identify who is who, talking about admixture...
The use of words as grotesque or pure in order to qualify—or to visually describe—admixture, are amongst other interesting aspects of this piece. But what is striking is that one top scientist/politician involved with the constitution of medical genomics; interprets the genomic boundary between the Mestizo and indigenous identity, following historical criteria, that goes against the fundamental logic of molecular anthropology and population genomics (cf. Gabriel et al. 2002; Zhu et al. 2004; Frazer et al. 2009).
In the previous segment, Dr. X is not only inverting, but using a basic shared knowledge of population genomics—Africans are older—in order to reinforce (consciously or not) the temporal divisions on which “Mestizaje” rests. As if the molecular anthropological stories of admixture (like the Mexican one) followed this pattern (i.e. Mestizos are younger and therefore they should have larger haplotype blocks). Approached through a syllogism the argument makes a lot of sense:
The counter intuitive story told by Haplotypes is hard to understand in and out of the laboratory. When discussing this “scientific curiosity” with Elisa—a laboratory technician at Illumina/Population genomics laboratory—she argued:
No...No Zapotecos are older, how can they be younger. It does not make sense, is endogamy, not temporality. I still think that the Zapotecos are older, than the Mestizos... it relates to endogamy, not time. They are older; they were here before the Mestizos.
The social-political template, on which Indigenous and Mestizo identities are built, definitely permeates the approach and understanding of population genomics. Discordances or “curiosities”, as the time story told by haplotypes, render visible the continuous efforts to make sense of population genomics, in not so new, and apparently, not so problematic terrains.Footnote 30 Another example of how time frames intervene in the interpretation of biogenetic identities can be found in Dr. Elias, population geneticist, explanation on how he differentiated between Indigenous and Mestizos individuals in his famous genetic studies: “well indigenous populations have been here for a very long time, before the Spaniards.... and the Mestizos are the new ones...”
Medical/population genomics is a young and very fragmented disciplinary field; as an example medical doctors working in genomics use the gen-ethnic labels without really knowing how they were produced. Public scientists and top officers inside the institute, as the ones I have interviewed for this article, are not really trained in population genomics. The much reduced number of specialists in population genomics and medicine in Mexico are the product of organisational efforts such as the INMEGEN, or are still to graduate from the BA in Genomic Science of Mexico’s National University (UNAM). Irma Silva-Zolezzi (human geneticist) knowledge about population genomics has been the result of an intense 5 year process of experimentation and autonomous learning in order to develop a map of common genetic variability in Mexico; rather than the product of formal training in the area of population genomics.
Once, that is said, it would be understandable that public figures in the field, make sense of knowledge gaps by making reference to popular and uncontested notions of Mexicanhood such as Mestizaje. Nevertheless, we think that the use of haplotypes and genetic variability, in order to reinforce what we already know about Mexican origins, reveals something much more interesting than the “ignorance” of notorious scientist in Mexico. What the confrontation between these two identities (popular vs. genomic) reveal, is the work of alignment or calibration that remains hidden, since it is re-articulation (Fullwiley 2008:617) or made to fit into century old notions of Mexican origins.
Popular assumptions of ethnicity deeply solidified in the imaginary of lay citizens, journalists and scientists alike, are not only deployed in order to sell the benefits of genomic science. The crisscrossing of national identity and biomolecular identities can be traced down to the very interpretation of genetic mapping. For STS scholars there might be no surprise in discovering that in the terrain of “neutral” DNA “Mextizaje” mediates the understanding of genetic groupings in Mexico; the same way it reveals race as—les Américains—know it in other cultural coordinates (Fullwiley 2008:706).
Unfortunately one of the aspects that these simplistic calibrations/re-articulations leave behind is the complexity and multiplicity involved in the emergence of this techno-scientific field. Making flat, what is really complex, and reproducing received knowledge, instead of problematising and exploring emergent properties and contradictions. Many of our informants would say such flattening is beneficial for science communication and political negotiations