Wednesday, September 30, 2015

Genetic Map of Mexico

Genetic map of Mexico
Map showing multiple autosomal genetic studies performed in the Mexican population

The map above is intended to be a map that  shows the European, Native American and African genetic admixture found in Mexico, maps like this one already exist, but what makes this one different are the next features

1 - This map aims to be the most complete and representative map ever done regarding the gentics of Mexico, the intention was to include studies for all the states but six were missing in the end

2 - Besides the classic circular graphs to represent the proportions of admixture, a color code for each  state  was implemented, with the colors meaning
Orange - representative sample
Blue - some kind of sample bias
Grey – No data

3 - The map prioritizes studies done on representative samples over studies with some sort of sample bias. A study is considered as representative when it declares to have been performed in the general population, with no sort, when it don’t claims to have any kind of sample bias, and when the results are consistent with other studies in the state that don’t have sample bias as well as being consistent with studies of the neighbouring states, it’s consistency with states all over the country is considered aswell, for example, if a study on Mexico city obtains an European admixture of 22% in average and then  there are studies in Oaxaca and Yucatan (states with very big percentage of native populations) on which the European admixture is of 30% and 40% sample bias is assumed because is impossible that Mexico City is more Amerindian than these states and is also inconsistent with studies done in Mexico City that have declare to be done with general representative samples (Studies that report a rather low European admixture in Mexico often declare to be done in diabetics, with diabetes being a disease that affects people with high native ancestry, thus there’s a sample bias).

4 - The sample bias category  includes studies done on diabetics (for the reasons stated above) asthma (asthma itself is correlated with low socioeconomic status which is related with higher indigenous ancestry in Mexico), studies that declare beforehand to have a sample bias (some studies are kind enough to state that their goal

List of studies

1 - Genetic admixture in three Mexican Mestizo populations based on D1S80 and HLA-DQA1 loci 
Study done in the three largest metropolitan areas of Mexico.
Study done in Mexicans from Nuevo Leon, San Luis Potosi and Zacatecas.

This study states to have been done using the most general population samples.
Study done in Tlaxcala and Coahuila, Saltillo appears blue because the study focuses on the transplanted tlaxcaltean population living in said city.
Study done in Mexicans from Nayarit, in the chart included Nayarit appears close to Monterrey, the study that is cited for Monterrey has an European average for Monterrey of 60%.
Study performed on Chihuahua, not many details about sample selection are given, but is inconsistent with neighbouring states so the state will be blue.

Study for Sinaloa, the African result appears rather high, specially compared with al the neighbouring states but Chihuahua, perhaps north African/semitic influence or sample bias? Not sure about it.

Study in north eastern Mexicans, the same sample was tested using two different methods.

Study done in Mexico city, Toluca, Queretaro, Morelos and  Puebla, the average for all is 52% European, 39% Amerindian and 9% African, to not put five circles with same value the admixtures were distributed based on other studies done in the region, making sure the total average ends up being 52%, 39% and 9%.

10 - Variation in the gene frequencies of three generations of humans from Monterrey, Nuevo León, Mexico.
An study done in Monterrey, claims to have been performed in 3 different groups (divided by generations) and there's four different results. It's not stated if the samples groups were related or not.

11 - A Genomewide Admixture Map for Latino Populations 
A genetic study of various Latin American populations, among them Mexicans from Mexico city.

12 - Gene frequencies and admixture estimates in a Mexico city population.
An old study done in Mexico city. Amerindian is the dominant admixture.

An old study from Mexico city, the European admixture is the dominant one.

An study performed in the autochtonous (indigenous) population of the state of Hidalgo. 
A study done in coastal cities of Mexico, the African and native american are considerably high and the study is old, but no more recent studies exist. Google Images seems to agree with them so it will be put as representative (orange color).

Study from 1996 done in cities from Oaxaca, Merida, Guanajuato and Coahuila, curiously Merida (in the State of Yucatan) is the city that obtained the highest European admixture despite that is one of the states with the highest percentage of Amerindians, Saltillo is again considerably  high in native american admixture and so is Leon. Both states, Coahuila and Guanajuato are bordered by states that have considerably high European admixture (Queretaro, San Luis Potosí, Jalisco, Nuevo Leon, Zacatecas) it’s quite a inconsistency and don‘t match with what I‘ve seen when I‘ve been to Torreón (although the Study is from Saltillo, which have transplanted Tlaxcaltecs on it), and google images, specially in the case of Coahuila seems to disagree. I haven’t been to Guanajuato so don’t know much about it. This study by the same author includes results of the principal study in this point plus other ones and reads as it‘s conclusion:  “the main ancestral contribution to non coastal lower middle class mestizo populations is Indian (above 50%)” this suggests a sample bias because economic stratification and similar admixture variations have been reported in lower class people in Mexico city (the definition of lower middle class itself varies a lot and the admixture degrees reported certainly match better with those of Mexicans of low income socioeconomic background rather than middle class, including middle low, this study satates that majoritarily Amerindian admixture has been a constantin studies done in low income Mexicans. The fact that the study is from the same author as the two other studies mentioned in this point do helps to sustain this theory.

Genetic structure of three Mexican populations based on the HLA-A system              
Another Study for Jalisco, in the graph included in the study the sample appears really close to the European samples, more than it is close to even  Nuevo Leon, but will be put as 73% to make it consistent, there is another study that found people from Guadalajara to be similar to Spaniards and white Brazilians but different from Amerindians, Black Brazilians and European Americans but because no percentages, charts or at least allele frequencies are accessible it will be omitted for now.

Study for Durango, European is lower than expected, probably because is a study by INMEGEN, which has sample bias as will be explained below, additionally this study is focussed in Amerindian diversity and genetic structure, so it wouldn’t have made sense to test  people with small or no visible Amerindian ancestry, Another couple of studies exist that have found the allelicfrequencies of Durango Mexicans to be very similar to whites, but no way of calculate the exact admixture is given.

The first study ever done by INMEGEN regarding genetic admixture, as opposed to it’s successors in this study the European admixture is dominant, even in Guerrero which is reported to have the least European Admixture at 51%, this study do test the same states that the study from 2009 test (number 20), this is Sonora, Veracruz, Zacatecas, Yucatan, Guerrero and Guanajuato, yet in the 2009 study the European admixture takes a drop of 16% in average, why? Many theories could be proposed to explain this, like the sample of the 2009 study  being 300 Mexicans and the sample of 2006 being 100, but if anything it would mean that they begun to select predominantly Amerindian Mexicans or as they say it in the "Material and methods" section of the 2009 study “self reported mestizos”. This methodology of selection is highly misleading: People at INMEGEN assume that, as reported in the census from 1921 the majority of Mexicans are mestizos (and that census in question didn't include the mestizo category, but one for mixed heritage, without specifying any degree). Another base INMEGEN could have used is that investigators have often opted for classifying Mexicans on a liguistic criteria that by lingüistic criteria, meaning that Mexicans that speak Spanish = mestizos regardless of how they look; and Mexicans who speak indigenous languages are considered amerindians; this leads to the country being 90% mestizos and 10% indigenous, which not only is incorrect from a biological/racial point of view, but also incorrect for the fact that no one has asked Mexicans what they identify as in nearly a century and nowadays the national census of the United States proves that the majority of Mexicans identify as white, so in reality they aren't testing 80-90% of the populatuion, they are testing less than half. That’s the only plausible scenario here regarding the increase of sample size, because have they kept expanding the sample size in a general manner at least one or two states would have go up in the European admixture, but instead all went down, although less notorious the African admixture went down aswell, specially in Veracruz and Guerrero. Another explanation could be that, as stated in the 2009 study the European markers used were Northern European, as opposed to Southern and Western European, that is from where the European ancestry of most Mexicans come from, (The largest European ancestries within Mexico are Spanish, French and Italian in that order), They don‘t use Siberians to calculate the native american ancestry in Mexicans so why would they use northern Europeans instead of western and southern ones?. To add more questions to the whole thing is the fact that in the webpage were the abstract of the 2006 study is aviable, the percentage of African admxture was changed from 5% to 10% some months ago after being 5% for almost 8 years, in fact, if look the phrase “A study conducted by Mexico's National Institute of Genomic Medicine (INMEGEN) reported that mestizo Mexicans are 58.96% European, 35.05% "Asian" (Amerindian mostly), and 5.03% African. Sonora shows the highest European contribution (70.63%) and Guerrero the lowest (51.98%) which also has the highest Asian contribution (37.17%). African contribution ranges from 2.8% in Sonora to 11.13% in Veracruz.” in google we find these percentages cited in many web forums and even images based on these numbers exist.

So why was it just changed? no one can know for certain what’s going on in INMEGEN, however as reported in this special investigation by Springer with interviews from insiders of INMEGEN that keep their identities as secret:conflicts of interest have took place.