Jump to content

Example of cancel culture...


Recommended Posts

I'm a British scientist and I’ve not only peer reviewed the original and found it to be false but I can easily prove it.

I'm more “related" to Cheddar Man than most people are.

I'm U5b1c (mitrochondrial haplotype), Cheddar Man is U5b1. I'm I1 and Cheddar Man is I2 (more distantly related by the Y chromosome haplotype). I have family not far from the area where Cheddar Man was found. Though I should mention, those haplotypes are actually quite weak in determining how related two people really are. The stupid thing is that when you determine being related like that then we run the risk of reaching the point of why not say everyone is related being in the same species or something.

main-qimg-0a5dbfd40be4e6ae3fa91b27666ed235
main-qimg-ca7c541855acd0f5d6dcf0eed223d43e
 
 
In human genetics, Haplogroup I (M170, P19, M258) is a Y-chromosome haplogroup . Haplogroup I (the letter I, not the number 1) can be found in most present-day European populations, most commonly in Scandinavia, Sardinia, and the Slavic & Bulgarian populations of the Western Balkans in southeastern Europe. Based on DNA dating analyses, Haplogroup I first arrived in Europe around 20,000-25,000 years ago from the Middle East, just prior to the onset of the last glacial maximum (LGM). Some speculate this migration was associated with the initial spread of the Gravettian culture [2] . Haplogroup I is closely related to Haplogroup J , most common in Semitic peoples; both Haplogroup I and Haplogroup J are descendants of Haplogroup IJ (S2, S22) . Haplogroup IJ is in turn derived from Haplogroup F that formed a founding middle east population 30-40kya. There are two principal subgroups of Haplogroup I (Y-DNA) : I-M253/I-M307/I-P30/I-P40 has highest frequency in Scandinavia, Iceland, and northwest Europe. I-S31 includes I-P37.2, common in the Balkans and Sardinia, and I-S23/I-S30/I-S32/I-S33, reaching its highest frequency along the northwest coast of continental Europe. [1] Rootsi et al. 2004 suggest that each of the ancestral populations now dominated by a particular subclade experienced an independent population expansion, believed to reflect different migrations of people during and immediately after the ice age. Haplogroup I Y-chromosomes have also been found among some populations of the Middle East, the Caucasus, and Central Asia, but they are found at frequencies exceeding 10% only among populations of Europe and Asia Minor, particularly among Germanic, Slavic, Uralic, and Turkic peoples, as well as among the Romance-speaking populations of France, Romania, Moldova, and Sardinia, the Albanian-speaking population of Albania, and the Greek-speaking population of Greece. Within Europe, several populations are distinguished by having a significantly lower frequency of Haplogroup I than the surrounding populations: these depressions in the frequency of Haplogroup I distinguish the populations of Italy and Switzerland from Germany and Sardinia, Iberia from southern France and Normandy, Greece from Albania and the Slavic peoples, and the Baltic Latvians from the Finnic Estonians. In all these areas, Haplogroup I populations are small relative to the dominant European haplogroups (R1b in Western Europe, R1a1 in Eastern Europe, and N in Northeastern Europe). Additional recommended knowledge Subgroups The subclades of Haplogroup I with their defining mutation, according to the 2006 ISOGG tree: I (M170, M258, P19) I* I1 (P38) I1* I1a (M253, M307, P30, P40) Typical of populations of Scandinavia and Northwest Europe, with a moderate distribution throughout Eastern Europe I1a* I1a1 (M227) Appears to be limited to a marginally low frequency of approximately 1% among Slavic and Uralic peoples of Eastern Europe; also detected in a single Lebanese man I1a2 (M21) I
 

Ancient DNA is not all that reliable. Cheddar Man has two reads, an older one came up as U5a:

https://isogg.org/wiki/Famous_DNA:Contested_DNA_Results#Cheddar_Man

Cheddar Man is also on gedmatch with two kits, SC8491738 and NW6414429.

I suspect these are from two different tests. Either way they do not match up entirely so it’s clear there’s likely some errors in reading. How truly precise each read is isn’t certain. We don’t have a perfect sample of his DNA for comparison to know what 100% really is. We can make some inferences from comparisons to other human genomes but this only goes so far. That alone throws analysis of his DNA into doubt.

You can actually download the genetic data they used from the Natural History Museum to determine Cheddar Man is black and then run it yourself.

https://www.nhm.ac.uk/content/dam/nhmwww/our-science/our-work/origins-evolution-futures/cheddar-man-pigmentation-data.xlsx

They used that with this tool:

 
 
HIrisPlex-S Eye, Hair and Skin Colour DNA Phenotyping Webtool 8px 9px 10px 11px 12px 13px 14px 15px 16px 17px 18px HIrisPlex-S Eye, Hair and Skin Colour DNA Phenotyping Webtool Welcome to the Department of Genetic Identification of Erasmus MC With the advancement of DNA phenotyping as a tool in Forensic and Anthropological usage, we now provide an easy to use interactive website to predict eye, hair and skin colour from DNA using the IrisPlex, HIrisPlex and HIrisPlex-S systems. This work is in collaboration with the Walsh laboratory of Indiana-University-Purdue-University-Indianapolis (IUPUI), USA. If you have genotypes from SEQUENCE DATA and wish to automate the conversion of genotypes into the correct allele format for prediction using this online tool, please see the Walsh lab tools page. Please see the HPS manual for instructions. We hope you enjoy using this tool and find it a useful addition to your analyses. If you have any issues regarding this website, please contact us at dnaphenotyping ... at ... gmail ... com * less than 1000 test runs were used to compute this metric due to lower numbers of this category within the database file upload example input file (header only, add your own data): hirisplexs.csv References Please cite all references when you use the HIrisPlex-S eye, hair and skin colour online prediction tool in your analyses. L. Chaitanya, K. Breslin, S. Zuñiga, L. Wirken, E. Pospiech, M. Kukla-Bartoszek, T. Sijen, P. de Knijff, F. Liu, W. Branicki, M. Kayser, S. Walsh. The HIrisPlex-S system for eye, hair and skin colour prediction from DNA: Introduction and forensic developmental validation. Forensic Science International Genetics https://doi.org/10.1016/j.fsigen.2018.04.004 S. Walsh, L. Chaitanya, K. Breslin, C. Muralidharan, A. Bronikowska, E. Pospiech, J. Koller, L. Kovatsi, A. Wollstein, W. Branicki, F. Liu, M. Kayser, Global skin colour prediction from DNA. Human Genetics, 2017. 136(7): p. 847-863. S. Walsh, L. Chaitanya, L. Clarisse, L. Wirken, J. Draus-Barini, L. Kovatsi, H. Maeda, T. Ishikawa, T. Sijen, P. de Knijff, W. Branicki, F. Liu, M. Kayser, Developmental validation of the HIrisPlex system: DNA-based eye and hair colour prediction for forensic and anthropological usage. Forensic Science International: Genetics. 2014 Mar;9:150-61. Funding support The initial research and development that led to this resource was supported in part by funding from the Netherlands Forensic Institute (NFI), the Erasmus MC University Medical Center Rotterdam, and by a grant from the Netherlands Genomics Initiative (NGI)/Netherlands Organization for Scientific Research (NWO) within the framework of the Forensic Genomics Consortium Netherlands (FGCN). Development and maintenance of this web resource is supported by Erasmus MC University Medical Center Rotterdam. The work of the Walsh laboratory is funded in part by the U.S. National Institute of Justice under grant number 2014-DN-BX-K031, the views presented here are for
 

When I put in Cheddar Man’s details I get this:

blue eye: 0.564

intermediate eye: 0.189

brown eye: 0.247

blond hair: 0.014

brown hair: 0.719

red hair: 0.011

black hair: 0.257

light hair: 1

dark hair: 0

very pale skin: 0

pale skin: 0

intermediate skin: 0.013

dark skin: 0

dark to black skin: 0.987

 
main-qimg-4b222f68b6330a5e4e97381bcfa8ca79

Now the first thing you should notice about this is some strangeness. Light hair but then only brown or black hair? No chance of dark skin? Did he have white hair like Julian Assange? Is that what the tool is saying? Lets say black with blue eyes and a white Afro? It’s like one of those characters you make in a game for a laugh.

The tool can’t even tell you if someone is just black. It can’t tell them apart from being dark to black.

Now to you this might look all very good but Cheddar Man isn’t living specimen. We can’t check him against the tool to see if it’s accurate.

So I ran myself against the tool. I’m very pale or quite pale. Everyone in my family is. I have very dark brown eyes which is also common in the family. I have fairly blond hair, a little dirty.

My DNA test is recent, through a fairly high quality service (fully sequenced). I have deep ancestry in the British Isles and then European, nothing outside for a very long time. Basically enough to say I’m pure British or pure European depending what time span you’re looking at. I’m a modern human which the tool should be better modelled for so what do I get?

blue eye: 0.949

intermediate eye: 0.035

brown eye: 0.016

blond hair: 0.732

brown hair: 0.234

red hair: 0.003

black hair: 0.032

light hair: 0.962

dark hair: 0.038

very pale skin: 0.036

pale skin: 0.044

intermediate skin: 0.894

dark skin: 0.006

dark to black skin: 0.02

 

Good luck even finding a photo of someone like that. I’m guessing the tool thinks I look like this:

main-qimg-ef1b2dcb5b61e585693227c520c65971

So according to the same tool they used for Cheddar Man, it says I’m half black, half white with blue eyes. This is totally wrong. The tool is barely better than guessing. The tool says I should more likely have blue eyes than Cheddar Man. I don’t.

This is what I actually look like:

main-qimg-cea4941b5c37399f5990318ae744048e-pjlq

The evidence is that the tool they used appears to be complete and utter trash. Scientists still haven’t entirely figured out exactly what genes lead to what skin colour.

The tool in question uses 1423 sample individuals. They say it uses 36 points for determining the skin colour for Cheddar Man. That’s 150094635296999140 possible permutations. Those SNPs are also likely only some of those involved so there’s ample scope for the tool to make some big mistakes.

I’ve tried to flip the input to see if they used the tool incorrectly but I still get junk results. Other genetic testing services can correctly predict my appearance much better.

23andme:

main-qimg-cee23bd52c1cbb4aed65ee6a3115b750
main-qimg-1593f2a51e8a7e3466df4e6dbae51d69
main-qimg-dd749c7dc7f737825f304fd755146e1f

23andme gets my hair wrong. Though it is correct on eye colour and skin colour. As you can see, these tools are woefully inaccurate still. Hirisplex is actually worse than 23andme. Even though 23andme gets my hair wrong, my hair still isn’t prohibitively improbable based on its prediction.

It would be interesting to see if I can get the raw data from one of those kits then upload it to 23andme. It still might not be reliable but it would be interesting to see what it produces.

Given the results from Hirisplex, assuming anything can be inferred from them at all, I’d say it’s most likely it gave Cheddar Man blue eyes when he didn’t have them just as it did for me.

I’d also say it most likely blackened him just as it blackened me. Though note, Hirisplex doesn’t actually make it exactly clear what it means by dark skin.

The evidence is basically that lack of evidence and that I’m relatively closely related to him with similar genes (we only have a few mutations differentiating us) and the tool scientists used mischaracterised me as being mixed skin colour as well as blue eyed when I’m not.

This is based on my real DNA. For example, here is my allele for rs1129038 on the HERC2 gene:

main-qimg-6f4ac71d7b9cbed76bfd7378d0faa25e

I’ve also double checked all the available SNPs on 23andme which also match:

main-qimg-67735e82688a57a30e8d1a2ada40c476

My full data combined with Cheddar Man’s is included (scientists always include their data otherwise it’s not science):

 
  • #,Gene,SNP,GRCh38,Allele,Forward,Flip,Me,Diff,CM,Diff,23andme,F0,F1,F2,F3,F4 
  • 1,MC1R,rs312262906,16:89919342,A,A,?,C,0,C,0,0,0,0,0,0,0 
  • 2,MC1R,rs11547464,16:89919683,A,A,?,G,0,AG,1,1,0/1,0/1,0/1,0/1,0/1 
  • 3,MC1R,rs885479,16:89919746,T/G>,A,A,G,0,G,0,1,0,0,2,0,0 
  • 4,MC1R,rs1805008,16:89919736,T,T,?,C,0,C,0,1,0,0,0,0,0 
  • 5,MC1R,rs1805005,16:89919436,T,T,?,G,0,G,0,1,0,0,0,0,0 
  • 6,MC1R,rs1805006,16:89919510,A,A,?,C,0,C,0,1,0,0,0,0,0 
  • 7,MC1R,rs1805007,16:89919709,T,T,A,C,0,C,0,1,0,0,0,0,0 
  • 8,TUBB3,rs1805009,16:89920138,C,C,?,G,0,G,0,0,0,0,0,0,0 
  • 9,MC1R,rs201326893,16:89919714,A,A,?,C,0,C,0,0,0,0,0,0,0 
  • 10,MC1R,rs2228479,16:89919532,A,A,?,G,0,G,0,1,0,0,0,0,0 
  • 11,MC1R,rs1110400,16:89919722,C,C,?,T,0,T,0,1,0,0,2,0,0 
  • 12,SLC45A2,rs28777,5:33958854,C>A,C,G,A,0,C,2,1,2/0,0/2,0/2,0,0/2 
  • 13,SLC45A2,rs16891982,5:33951588,C>G,C,G,G,0,C,2,1,2/0,0/2,0/2,0,0/2 
  • 14,KITLG,rs12821256,12:88934558,G,C,C,T,0,T,0,1,0,0,2,0,0 
  • 15,LOC105374875,rs4959270,6:457748,A,A,T,CA,1,CA,1,1,1,1,1,0,1 
  • 16,IRF4,rs12203592,6:396321,T,T,?,C,0,T,2,1,0/2,0/2,0/2,0/2,0/2 
  • 17,TYR,rs1042602,11:89178528,T/C>A,A,A,CA,0,C,0,1,1/0,1/0,1/0,1/0,1/0 
  • 18,OCA2,rs1800407,15:27985172,A/C>,T,T,C,0,C,0,1,0,0,0,0,0 
  • 19,SLC24A4,rs2402130,14:92334859,G>A,G,C,A,0,A,0,1,2,0,0,0,0 
  • 20,HERC2,rs12913832,15:28120472,T/A>G,A,A,GA,0,G,0,1,1/0,1/0,1/0,1/0,1/0 
  • 21,PIGU,rs2378249,20:34630286,C,G,?,GA,0,A,0,1,0,0,0,0,1/0 
  • 22,LOC105370627,rs12896399,14:92307319,T,T,A,T,2,G,0,1,2/0,2/0,2/0,0,2/0 
  • 23,TYR,rs1393350,11:89277878,T/G>A,A,A,GA,0,G,0,1,0,0,0,1/0,1/0 
  • 24,TYRP1,rs683,9:12709305,G/C>,C,!,C,0,C,0,1,0,0,0,2,2 
  • 25,ANKRD11,rs3114908,16:89317317,T>C,T,A,C,0,CT,1,0,2/1,0/1,2/1,0,0/1 
  • 26,OCA2,rs1800414,15:27951891,C,C,?,T,0,T,0,1,0,0,0,0,0 
  • 27,BNC2,rs10756819,9:16858086,G>A,G,C,A,0,?,?,1,2/?,0/?,2/?,0/?,0/? 
  • 28,HERC2,rs2238289,15:28208069,C/A>G,G,G,GA,0,A,0,1,1/0,1/0,1/0,1/0,1/0 
  • 29,SLC24A4,rs17128291,14:92416482,C/A>,G,G,A,0,A,0,1,0,0,0,0,0 
  • 30,HERC2,rs6497292,15:28251049,C/A>G,G,G,GA,0,A,0,0,1/0,1/0,1/0,1/0,1/0 
  • 31,HERC2,rs1129038,15:28111713,G/C>T,C,!,CT,0,T,0,1,1/0,1/0,1/0,1/0,1/0 
  • 32,HERC2,rs1667394,15:28285036,C>T,C,!,CT,1,T,0,1,1/2,1/0,1/0,0/0,1/0 
  • 33,TYR,rs1126809,11:89284793,A,A,?,GA,1,?,?,0,1/?,1/?,1/?,1/?,1/? 
  • 34,OCA2,rs1470608,15:28042975,A/G>T,T,T,GT,0,G,0,1,1/0,1/0,1/0,1/0,1/0 
  • 35,SLC24A5,rs1426654,15:48134287,G,G,?,A,0,G,2,1,0/2,0/2,0/2,0/2,0/2 
  • 36,ASIP,rs6119471,20:34197406,C>,G,G,C,2,C,2,0,0,0,0,0,0 
  • 37,OCA2,rs1545397,15:27942626,T,T,?,A,0,A,0,0,0,0,0,0,0 
  • 38,RALY,rs6059655,20:34077942,T/A>G,A,!,GA,0,G,0,0,1/2,1/2,1/2,1/0,1/0 
  • 39,OCA2,rs12441727,15:28026629,A,A,?,GA,1,G,0,0,1/0,1/0,1/0,1/0,1/0 
  • 40,MC1R,rs3212355,16:89917970,A/C>T,T,T,CT,0,?,?,0,1/?,1/?,1/?,1/0,1/? 
  • 41,DEF8,rs8051733,16:89957798,C/A>,G,G,A,0,A,0,0,0,0,0,0,0 

An interesting question with this is why am I more genetically diverse than Cheddar Man? I can speculate some possible explanations but Cheddar Man’s lack of genetic variation in this read out does raise some questions as to the integrity of the data. I’d first wonder if he might have been inbred.

Something else that puzzles me is that there are no multiple alleles with their duplicates. Never AT or GC. Is this a limitation in sequencing? It can’t be a coincidence for mine. Perhaps for Cheddar Man, but it appears consistent.

If you look closely, we’re deciding Cheddar Man’s appearance as different to mine based on only seven alleles that are unique to each of our respective genetic one man pools.

  • rs28777 (A -> C) [darker,east-asian,modern]
  • rs16891982 (G -> C) [darker,east-asian,modern]
  • rs12203592 (C -> T) [lighter,european,modern]
  • rs12896399 (T -> G) [insignificant]
  • rs1426654 (A -> G) [darker,east-asian]
  • rs11547464 (G -> A) [lighter,european,modern,half]
  • rs3114908 (C -> T) [half,insignifcant,global,modern]

This is a tiny sample size. I suspect if I included my parents one or two of these unique alleles might appear. It’s only the three SLC SNPs that the tool is really using to put him potentially in the black category. That’s not a huge amount to go on.

I’d argue the only SNP that Cheddar Man has that I can find data for not being present in the modern population is rs1426654. As for the rest I don’t know their proportions by location. rs1426654 is still found across Eurasia/America in non-black (though darker) populations.

Based on available literature the distribution of rs16891982 is similar to that of rs1426654. There’s less data on rs28777 but one data set does shows the allele as significantly present in parts of China. the odd thing about rs28777 is that it seems to have been barely mentioned outside of HIrisPlex-S literature.

At a glance of these global frequencies then HIrisPlex-S would probable make similar mistakes with a fair proportion of people with Central/East Asian ancestries as well as people from a few other niches. Though it might take a few hundred individuals to get a good few with all those alleles, you should probably find the tool sometimes says they’re black when not. Otherwise I’m not hearing about these few percent or one in a thousand black Chinese, etc.

Though Cheddar Man’s genes are stated as ancestral, from what I can tell, comparing to modern frequencies, on these SNPs he’s basically within the modern range for Eurasian combinations though on the more slightly unusual side. Basically, he actually looks like a mix of East Asian and European rather than African and European. That or it’s ambiguous. I’ve never seen a mix of East Asian and European ever come out as anything but something between how those two look. What this tool seems to be saying is that Europeans and East Asians mixing should fairly regularly produce black people!

Assuming the data I’m looking at is reliable then what this tool is actually saying is that interracial offspring between East and West Eurasians should be warned that they have a fair chance to produce a black child and to not assume that the wife has been unfaithful. I guess realistically it would want to take a few generations for the homozygotes to start popping up and coming out in force so if we look at some places where the two have mixed a lot for a while what does that look like?

Compared to my DNA I would anticipate that Cheddar Man is more than 90% within the modern pool for the SNPs surveyed in his pigmentation mutations if we were to compare him to a larger sampling. The thing is, what this tool seems to be saying is that we would have really common instances of sort of the opposite of albinism. Random occurrences of parents producing black children would be fairly well heard of. I’ve never heard of this. It would be one of the first things you would heard about scientists having isolated the mutations for.

My intuition is that the only gene set missing from Cheddar Man may turn out to be the few hyper white mutations, though this could be wrong as well because he may have had some variants for whiteness that were less favourable to selection than the modern versions (including in combination).

Starting at that point, the overall trend is for the SNPs to send him in the direction of being lighter. Though it’s only two of them that really do that which isn’t a lot. Whatever values I use for the missing SNPs makes him more black which seems suspect to me.

The SNP on the ASIP gene appears to be significant in skewing it toward darker. It’s possible that it might be sometimes associating the probability of certain SNPs twice over.

I found mention of the ASIP SNP being the wrong way around in the manual. The Cheddar Man analysis doesn’t take this into account as far as I can tell. One of the SNPs that’s missing for Cheddar Man is also mentioned as especially important for determining skin colour.

Though it appears to be the SNP rs1426654 alone that makes black skin highly probable. That’s a tiny mutation. Again, the tool appears to be using rs12913832 as the sole decider of if you have brown eyes otherwise the default is blue.

You can see straight away that something is strange with rs12913832.

Geno Mag Summary (A;A) 0 brown eye color, 80% of the time (A;G) brown eye color (G;G) 2.5 blue eye color, 99% of the time rs12913832 is a SNP near the OCA2 gene that may be functionally linked to blue or brown eye color , due to a lowering of promoter activity of the OCA2 gene. Blue eye color is associated with the rs12913832 (G;G) genotype.[ PMID 18172690 , PMID 18252222 ] For green versus blue eye color rs12913832 in OCA2/HERC2 has a score of 51.5 and an estimated allelic OR of 8.43 . The SNP rs1667394 in this same region has an estimated OR of (4.85–10.06). 10.1371/journal.pgen.1000993 rs12913832 is also part of a haplotype spanning 166kB on chromosome 15, defined by 13 SNPs listed below, that is found in 97% of all Caucasians with blue eyes. In this haplotype, variations in rs1129038 and rs12913832 are relatively common in Caucasians though rare among other racial groups.[ PMID 18172690 ] [ PMID 18650849 ] rs12913832-T (brown eye) homozygotes compared to rs12913832-C (blue eye). correlations with skin, eye, and hair color variation. The "h-1" haplotype found in homozygous state in 97% of individuals with blue eye color is composed as follows [ PMID 18172690 ]: rs4778241 (C) rs1129038 (A) rs12593929 (A) rs12913832 (G) rs7183877 (C) rs3935591 (G) rs7170852 (A) rs2238289 (T) rs3940272 (C) rs8028689 (T) rs2240203 (A) rs11631797 (G) rs916977 (G) blog coverage 10.1038/ncomms10815 A genome-wide association scan in admixed Latin Americans identifies loci influencing facial and scalp hair features [ PMID 19278018 ] Brief communication: Blue eyes in lemurs and humans: Same phenotype, different genetic mechanism [ PMID 19208107 ] Interactions between HERC2, OCA2 and MC1R may influence human pigmentation phenotype. OMIM 227220 Desc SKIN/HAIR/EYE PIGMENTATION, VARIATION IN, 1; SHEP1 Variant Related also OMIM 605837 Desc HECT DOMAIN AND RCC1-LIKE DOMAIN 2; HERC2 Variant Related also GWAS snp PMID [ PMID 20463881 ] Trait Eye color traits Title Digital quantification of human eye color highlights genetic association of three new loci Risk Allele P-val 0 Odds Ratio None None GWAS snp PMID [ PMID 20585627 ] Trait Eye color Title Web-based, participant-driven studies yield novel genetic associations for common traits Risk Allele A P-val 3E-52 Odds Ratio 8.43 [NR] [ PMID 20457063 ] Human eye colour and HERC2, OCA2 and MATP [ PMID 22615734 ] The Classical Pink-Eyed Dilution Mutation Affects Angiogenic Responsiveness [ PMID 22234890 ] HERC2 rs12913832 modulates human pigmentation by attenuating chromatin-loop formation between a long-range enhancer and the OCA2 promoter. [ PMID 19340012 ] Genome-wide association study of tanning phenotype in a population of European ancestry. [ PMID 19472299 ] Genotyping of five single nucleotide polymorphisms in the OCA2 and HERC2 genes associated with blue-brown eye color in the Japanese population. [ PMID 19619260 ] Genetics of human iris colour and patterns. [ PMID 19668368 ] Ancestry analysis in the 11-M Madrid bomb attack i
 

Note, that GG always leads to blue eyes is disputed. This may only apply for the modern genomes tested.

The question is: Why is the tool asking for T?

main-qimg-616f8897b1407a926f3da256b1c9ec7b

T doesn’t seem to be an accepted allele. However, if you put A in instead, it seems to make sense. I get this when I correct that that mistake in the tool:

blue eye: 0.184

intermediate eye: 0.15

brown eye: 0.665

 

So if that is wrong, what else is wrong with it? I could potentially find out, but that’s not my job. The tool is a bust.

That said, I think the ASIP 2 gene has a typo as well but a more subtle one. Usually you put the most frequent allele as the default then put the count of the rarer allele. The ASIP 2 SNP however might be the common one if the precedence convention is correct on NCBI: C>A,G. When I correct for that I get more sensible results again for both individuals. Still not perfect for me but an improvement so probably another bug down. I wonder if they messed that up because the tool doesn’t support multiple multiple rare alleles for each SNP?

This looks like a typo to me. I had to check everything two or three times if not more. Went through the effort to check for my typos only to find it was theirs. I’m also just doing this casually. They’re doing it professionally. It looks sloppy to me.

Another glaring issue is that they’re going to keep finding more mutations that have an effect. I’m not sure but I think they may have added one or two more SNPs since Cheddar Man was first tested on it. It could turn out to be hundreds or even thousands of SNPs that have various effects.

There are three missing SNPs for Cheddar Man. It’s possible they had these but didn’t publish them. I would guess he didn’t have those alleles, being the rarer ones and Cheddar Man has more often the less rare allele than the rare one relative to whatever the original dataset is or criteria.

There’s a question of what exactly is the underlying model doing when putting the most common for all of the missing ones makes him darker with the rarer ones making him darker still.

A fun game to play might be to go through all the SNPs. See which mutations make him darker. See what proportion those are. Then ideally a point of comparison would be wanted to see which direction evolution was going in. However, the tool is too useless for that kind of scientific pursuit.

There are probably SNPs that either individual should have checked but aren’t due to errors in the tool as well. I’d need to brute force combinations and look at those that accurately match either myself or 23andme. If I find one that should be on when it’s not, then it may also apply to Cheddar Man.

Performing a brief check, rs885479 looks like it’s in error as well looking for a T when GAC are the only reported variants. The errors in the tool are so frequent that not only is is largely useless but it looks like deliberate sabotage.

I’ve attempted some prospective fixes which appear to improve the results fairly well. In doing so various biases in the tool have presented. It’s very strongly biased toward intermediate skin and difficult to steer it away from that without also invoking a strong bias for red hair.

When I fix the tool for myself, Cheddar Man moves in the direction of more modern Europeans though not all the way. It also presents a more plausible character. The tool does however show signs of being somewhat confused still. G is under represented in the set which is a statistical suggestion of mistakes.


After further examination I have gotten to the bottom of why it appears as though someone has encrypted the input to this tool with an XOR making it hard to reliably use. It’s similar to the endian problem in programming.

 
 
By standard convention, the locations of DNA variations (SNPs) are based on their chromosomal position. This number changes every time a new reference human genome assembly is released; the current "build" is GRCh38, and the previous one, which is still used by many sources for a variety of reasons, is GRCh37 (also known as hg19). Genes are read (transcribed) in either the forward or reverse direction, as numbered along the chromosome. If a new build comes along that flips a large segment of a chromosome, the gene direction will change. As a result, at different times, as well as in different publications or different databases, the same SNP can be defined as being on the forward (plus) or as being on the reverse (minus) strand. In terms of the nucleotides for that SNP, the pairing of A with T,and C with G, in the DNA double helix means that an A on the plus strand by definition is a T on the minus strand, and vice versa, and a C on the plus strand means a G on the minus strand (and vice versa). The way a SNP is defined is based on its flanking sequences, so barring processing errors at dbSNP, the major and minor alleles of a SNP should not change between builds, however, the position number along the chromosome will change, and whether it is on the plus or minus strand may change. SNPedia currently deals with these complexities using two fields, Orientation , and, StabilizedOrientation , which are usually seen on each SNP (rs#)page at the top of the sidebar on the right of the page. Orientation indicates the orientation reported in the most current build (currently GRCh38), which is reported below the genotypes in the Reference field. The chromosome number and nucleotide position in that reference build is shown next. StabilizedOrientation is the orientation that is relevant to the genotypes that have been defined in SNPedia for each SNP, allowing Promethease to minimize confusion despite how reference builds may change over time or how companies may report genotypes in a person's raw data. [Both fields are carefully controlled and you should not edit them; if you believe either is incorrect for a given entry, report it to us and we will investigate.] Companies follow their own protocols, which are often different. 23andMe currently reports all genotype data based on the plus strand of GRCh37, whether or not dbSNP defined the SNP as being on the plus strand (in that build or any other build); they explain this for their customers here , here and here . This often leads to confusion since a 23andMe customer may see a genotype in their raw data that will not match the genotypes defined by dbSNP, in SNPedia, or in their Promethease report. This will most often happen when the StabilizedOrientation is minus. In these cases, the alleles need to be "flipped" to match: This can still lead to some ambiguous flips . In a Promethease report, this is done automatically as Promethease correctly flips genotypes from each company's orientation into the orient
 

https://www.ncbi.nlm.nih.gov/core/assets/snp/docs/RefSNP_orientation_updates.pdf

Basically, scientists can’t make their minds up about how to represent genetic data and which strand to use. Perhaps the tool uses an older convention or something and is out of date. The effect is that it’s near impossible to properly use it.

This problem goes beyond this tool. It’s a huge problem in genetics where bits are getting flipped. It’s very hard to in any situation be entirely sure which way around things are. In any circumstance, either of A, C, G or T are useless on their own. It needs to be signed. -A or +A for example. The missing sign in documentation and literature is a case of loss of precision.

If something isn’t versioned, unless you can deduce it from the data then it becomes garbage if it fails to disclose orientation. You now also need some decryption or conversion tool from one version to the other.

Even without this, the tool in its manual shows quite high error rates in testing (10% or higher). I can probably fix a lot of the input based on this but it’s tedious work. the manual says the SNPs other than ASIP are in the correct orientation. This statement is opinionated.

Correcting for the ASIP gene orientation issue in the data gives Cheddar Man as intermediate skin, blue eyes and brown hair. A somewhat unlikely arrangement but a more likely one than before. In particular, it’s in keeping with what we might expect from populations in Eurasia and the trajectory of evolution toward lighter pigmentation over time in Europe.

The question we really have to ask is how often is this mistake happening in genetics? I can’t see that it is humanly possible to avoid this mistake a lot especially when relying on other people’s work.

I’ve had similar problems in collaboration like this myself. Some people have their x, y and z axis oriented differently by default. Some people have degrees orientated like a block with 0 at the top. Other’s have it oriented sideways like a protractor with 0 on the left.

rs885479 has clearly flipped from A to T. If it had flipped from G to C then we’d be in trouble. We’d be able to say flipping hell. After seeing this I’m going to treat all findings involving genetics, especially alleles with a lot more scepticism than I usually do.

The question I have is why hasn’t this come up before? Anyone working in genomics regularly would have to know about this pitfall. Once you know about it you can fix some alleles in the tool on sight but others are ambiguous. There’s a double amount of confusion from which is the most frequent or base allele.

Even if you’e aware of it I’m not sure it’s possible to fully fix the tool from a user perspective. Not without a large sample set and access to the impact of all the alleles it surveys. The blunders this must be causing in science and the data loss is probably enormous. I’d call it a disaster. There are probably hundreds of papers that can simply go into the bin.

What order are my alleles in? What about Cheddar Man’s? See how much larger this problem is. All the alleles are in question. Just from this post alone you can see we have a huge reproducability problem. I suspect the people who made the tool can get it to work properly when they’re sequencing the DNA themselves. Anyone else trying to us it is up the creek without a paddle.

Finally, to settle this once and for all, I have found that the vendor has else where released a fix or a remap. This goes toward confirming many of the flips and errors I have noticed including the mistake in the ASIP gene turning Cheddar Man black:

online$rs6119471_C<-gsub('G/G', '2', online$rs6119471_C)

 

The limitation in this confirmation is that the allele could be double flipped. This is the mess that we’re in now. The script also confirms the flip of rs885479. This data should be on the page with the tool. Not hidden away elsewhere. The Cheddar Man data also does not include orientation.

I can’t include mine though it’s supposed to be all forward orientated. The orientation not being stated unequivocally is something I’d have to take up with nebula genomics. Though I see no obvious signs of flips in mine against the alleles from a source that always used forward orientation. The orientation of Cheddar Man’s alleles looks forward as well.

After correcting for this and running my results, it still darkens me though somewhat more modestly. Either I’ve made a typo or there’s a problem beyond remapping the alleles. The tool seems to think I’m at least a little bit shady like this guy:

main-qimg-f8f2a9c71548ec1e7bbab4e4ea3d41a4

Applying to corrections also gives Cheddar Man intermediate skin:

intermediate skin: 0.752

dark to black skin: 0.248

 

This needs to be double checked but it makes more sense. I also think this is in the realm where the tool is confused. It probably can’t reliably model these permutations.

The manual says these results can be interpreted as between intermediate and dark. By dark it means this:

main-qimg-ba7a05f5a8d12b50394d7ea8084c4891

https://hirisplex.erasmusmc.nl/pdf/hirisplex.erasmusmc.nl.pdf

Not this:

main-qimg-749c6fcc70a9d2d3d0cecf395f6cb8b1

If we’re going to be artistic then at least do it with style, the real Cheddar Man:

main-qimg-f78e02810b1b5a41f70814ebceb8e061

His hair just falls short of their guide for interpretation. It seems to suggest light brown but the presence of the variable for black hair would potentially suggest more intermediate. Though I believe you’re just meant to ignore the secondary result and dark or light for black hair.

The criteria for eye colour is so simple that it makes me wonder. Some results may fall into an undefined category with no explanation how to interpret them other than unknown.

The presence of rs11547464 may be an indicator of the direction evolution was taking his pigmentation. The same for rs12203592 which is apparently fairly well confined to small regions. With a large margin of error what we actually get is this:

main-qimg-34fcabf93b86c5d912bc9101cf93ed34

Perhaps not unlike this:

main-qimg-6a84b90aedefe8a7d716f8b374f201ca.webp

I have however been able to establish that there’s scientific evidence that rs12913832 doesn’t cause blue eyes in East Asians. This creates valid scope for uncertainty about Cheddar Man’s eye colour.

I’m not the first person to discover this mistake it turns out.

I am not sure how he got pale skin to show up however.

I also have rs683 which shifts me ever so slightly toward being darker, almost insignificantly. It has a very large effect on Cheddar Man though. It doesn’t have a large enough effect on my results to be able to find it as suspect as the ASIP variant was. However, there is a lack of public data on the SNP to be able to verify it making it suspect.

Playing a game of making possible recombination of someone from Cheddar Man’s pool shows a huge exaggeration toward black for many genes that without the SLC modifications you don’t get. rs3212355 and rs1126809 each cause a really big shift and I’ve got one of each of those. This large shift to black for any SNP slightly favouring darker skin is fairly consistent which smells fishy to me. It’s on the edge of not knowing if he was black or not and guessing.

The tool is saying the gene pool he is in ranged from black to intermediate. Perhaps even some redheads. If we consider the bias toward intermediate then it would be pale to black which is a fairly large range. It’s almost stuck on intermediate for me so could be for Cheddar Man too.

For many of the SNPs where I have two, that means I have one that makes me darker in many cases. These are mutations I have that would make me darker and Cheddar Man has the versions for being lighter.

Four of the alleles I have that Cheddar Man doesn’t have but I do make me lighter when I have them and six make me darker. However, some are such a small difference that it’s most likely no difference and just noise. I end up with something that looks like two and three.

For a small sample size, this difference isn’t significant. It’s practically 50/50 which way it’s going. It’s curious that the gene for brown eyes makes me a very tiny amount lighter. For Cheddar Man, it turns him black.

I don’t think these SNPs really come into play for me when it comes to making me darker or lighter. There’s a good chance the same would apply to Cheddar Man. It might be fair to assume that earlier evolution of lighter pigmentation had more instability, like this tool.

Of my variable alleles I only have to change three to get this:

pale skin: 0.066

intermediate skin: 0.374

dark skin: 0.545

dark to black skin: 0.016

 

I don’t have to change very many to get 95% dark skin. On face value it seems a little too improbable. Though I’d need to check the frequency of those alleles and features in the population at large.

Perhaps this can provide some insights:

https://www.researchgate.net/publication/316918714_Global_skin_colour_prediction_from_DNA/link/5f74833fa6fdcc008648f1a1/download

Possible other suspect broken alleles could perhaps be determined from the table in this.


I believe the interpretation of rs1426654 may be mistaken. There’s this “one simple genetic trick” to turn African skin into European skin mentality and I don’t think that’s accurate. Evolution surely had other mutations to favour that impact pigmentation before stumbling upon that one.

Evolution works in increments and the notion that it did nothing to select for other polymorphisms sure to be available for more than 40000 years seems bonkers. You can see three available polymorphisms likely to contribute to pigmentation in Cheddar Man’s DNA alone that evolution can select surely against.

 
 
Bertolotti, A., Lasseaux, E., Plaisant, C., Trimouille, A., Morice-Picard, F., Rooryck, C., Lacombe, D., Couppie, P., Arveiler, B. Identification of a homozygous mutation of SLC24A5 (OCA6) in two patients with oculocutaneous albinism from French Guiana. (Letter) Pigment Cell Melanoma Res. 29: 104-106, 2015. [PubMed: 26491832] [Full Text: https://doi.org/10.1111/pcmr.12425] Crawford, N. G., Kelly, D. E., Hansen, M. E. B., Beltrame, M. H., Fan, S., Bowman, S. L., Jewett, E., Ranciaro, A., Thompson, S., Lo, Y., Pfeifer, S. P., Jensen, J. D., and 36 others. Loci associated with skin pigmentation identified in African populations. Science 358: eaan8433, 2017. Note: Electronic Article. Erratum: Science 367: eaba7178, 2020. [PubMed: 29025994] [Full Text: https:/www.science.org/doi/10.1126/science.aan8433?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%3dpubmed] Lamason, R. L., Mohideen, M.-A. P. K., Mest, J. R., Wong, A. C., Norton, H. L., Aros, M. C., Jurynec, M. J., Mao, X., Humphreville, V. R., Humbert, J. E., Sinha, S., Moore, J. L., and 13 others. SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans. Science 310: 1782-1786, 2005. [PubMed: 16357253] [Full Text: https:/www.science.org/doi/10.1126/science.1116238?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%3dpubmed] Mack, M., Kowalski, E., Grahn, R., Bras, D., Penedo, M. C. T., Bellone, R. Two variants in SLC24A5 are associated with 'tiger-eye' iris pigmentation in Puerto Rican Paso Fino horses. G3 (Bethesda) 7: 2799-2806, 2017. [PubMed: 28655738] [Full Text: https://dx.doi.org/10.1534/g3.117.043786] Mondal, M., Sengupta, M., Samanta, S., Sil, A., Ray, K. Molecular basis of albinism
 
 
 
Protein Sodium/potassium/calcium exchanger 5 ( NCKX5 ), also known as solute carrier family 24 member 5 ( SLC24A5 ), is a protein that in humans is encoded by the SLC24A5 gene that has a major influence on natural skin colour variation. [5] The NCKX5 protein is a member of the potassium-dependent sodium/calcium exchanger family. Sequence variation in the SLC24A5 gene, particularly a non-synonymous SNP changing the amino acid at position 111 in NCKX5 from alanine to threonine , has been associated with differences in skin pigmentation . [6] The SLC24A5 gene's derived threonine or Ala111Thr allele (rs1426654 [7] ) has been shown to be a major factor in the light skin tone of Europeans compared to Sub-Saharan Africans , and is believed to represent as much as 25–40% of the average skin tone difference between Europeans and West Africans . [5] [8] Possibly originating as long as 19,000 years ago, it has been the subject of selection in the ancestors of Europeans as recently as within the last 5,000 years, [9] and is fixed in modern European populations. [10] [11] [12] It was introduced into Khoisan people via "back-to-Africa" migration around 2,000 years ago is partly responsible for their differing skin tone to most other African populations. [13] In human, the SLC24A5 gene is located on the long (q) arm of chromosome 15 at position 21.1. The SLC24A5 gene, in humans, is located on the long (q) arm of chromosome 15 on position 21.1, from base pair 46,200,461 to base pair 46,221,881. [5] Protein [ edit ] NCKX5 is 43 k Da protein that is partially localized to the trans-Golgi network in melanocytes . Removal of the NCKX5 protein disrupts melanogenesis in human and mouse melanocytes, causing a significant reduction in melanin pigment production. Site-directed mutagenesis corresponding to a non-synonymous single nucleotide polymorphism in SLC24A5 alters a residue in NCKX5 (A111T) that is important for NCKX5 sodium-calcium exchanger activity. [6] Effect on skin color [ edit ] Global frequency distribution of the SLC24A5 gene's ancestral Ala111 allele (yellow) and its derived Ala111Thr allele (blue). SLC24A5 appears to have played a key role in the evolution of light skin in humans of European ancestry. The gene's function in pigmentation was discovered in zebrafish as a result of the positional cloning of the gene responsible for the "golden" variety of this common pet store fish. Evidence in the International HapMap Project database of genetic variation in human populations showed that Europeans, represented by the "CEU" population, had two primary alleles differing by only one nucleotide , changing the 111th amino acid from alanine to threonine , abbreviated " A111T ". [5] [14] [15] The derived threonine allele ( Ala111Thr ; also known as A111T or Thr111 ) represented 98.7 to 100% of the alleles in European samples, while the ancestral or alanine form was found in 93 to 100% of samples of Sub-Saharan Africans, East Asians and Indigenous Americans. Th
 

If you used that SNP to assess someone from Japan then it would say that they’re black as in that they look like a sub-Saharan African. You see the problem? Though it might not give us all the answers, I’d like to know what happens if you take a white European genome and give them only the African allele for this SNP.

From what I understand this allele is also all over the Mediterranean, Middle East, India, Asia, etc where they have darker skin so to say it’s that alone that makes us white is curious indeed.

Cheddar Man may have actually been yellow! Perhaps more orange. Though there is a rumour that the blue eye SNP causes white skin instead in orientals. Many of these SNPs aren’t tested for other populations or in combination with other mutations.

There’s an over simplistic assumption that lighter skin only evolved once in Europe and the surrounding region.

The optical qualities of skin show a great variation under selective pressure with convergent evolution not uncommon. Things like blue eyes may be a remnant or after effect of divergent convergent evolution that then merged. This is particularly likely in a region where the selection is much higher.

The pattern in Europe looks potentially like mixing between multiple depigmentation genes in groups semi isolated, enough for evolution to take different paths toward the same goal but not enough for it to completely eliminating flow leading to selections toward both gene sets settling on an intermediate range of coexisting mutations.

There is a belief that everyone with the rs12913832 allele for blue eyes has a single ancestor. That’s not necessarily how evolution works. It’s a single mutation, it may have occurred many times including in Africa over all this time. The mutation was probably only conserved where it offered a benefit.

If that’s not the case, what are the chances the mutation never occurred in anyone in Africa over the tens of thousands of years or if it did then why wasn’t it sexually selected for as proposed is the case in Europe?

The blue eye gene alone actually really confers no obvious advantage. Typically, we have to invent advantages such as sexual selection. The gene involved is not only associated with eye colour but reportedly has a complex relationship with other factors such as tanning and pigmentation itself. There’s a good chance that blue eyes is actually a byproduct.

Tanning is really under appreciated. The evolutionary pressure on people in Eurasia is to have low melamine by default then produce it on demand. The pressure on black people appears to more so be to have it always on by default.

If I were generically engineering humans, functionally what would normally make the most sense is to be white by default with very little hair. Then grow it or release it very accurately in response to the environment rather than it having to evolve over generations to keep up with environmental changes. In my speculation of the merger of different genetic strategies, this also involves perhaps some super effective tanners and others with always on low skin pigmentation. When you have to put on sun block appropriately, that’s a point evolution is trying to reach, it just hasn’t at this point. People stuck as black or white or rather with more or less UV block isn’t actually ideal and we’d genetically engineer ourselves differently.

I strongly suspect blue eyes is an after effect. It’s probably secondary rather than primary. It actually looks like a genetic left over from two merging genetic strategies for depigmentation that results in unnecessary or arbitrary mutations. Blue eyes may not have been present in the original form where the mutations were present. Instead it being an arbitrary side effect of combining two different genetic strategies that achieve the same purpose.

Some combinations would probably have been harmful and selected against eliminating a full copy of each original commonly available in the present population. Some harmful combinations like that may occur but not be recognised as modern amenities and knowledge avoids the harm that would occur in nature. If you consider the variation in my genetics, it’s like I’m mixed between Cheddar Man and some other strain of human for genes pertaining to pigmentation.

Could this be the source of my greater genetic diversity for these SNPs? Could Cheddar Man’s unique SNPs be pathological and the dual shared ones not so? Unfortunately, a mere individual or two is a limited sample size and there are other explanations.

People also forget that the skin is dynamic nor the complex ways in which evolution works. For all we know, the selective pressure for skin to lighten may have selected for albinos. However this is a case of too light so is then selected in the other direction. There are a lot of complex scenarios involved.

I’m also certain that there’s some case of people mistaking white and non-white. There are modern European genes for being hyper-white. Some people seem to think that anything other than that must be therefore modern African.

The practice of exhuming a population’s ancestors and then making displays of them that is not only based on poor science but used for propaganda purposes should be banned. Cut the funding and ban the scientific discipline, with artefacts only allowed to be placed in cold storage until the industry is mature enough to do it properly.

There’s clearly a lack of licensing in place and no auditing. This must change. We have it in other industries so why not these ones?


One of the issues raised by new scientist behind a paywall (not accessible to the general public or basically a non-publication), is that the tool is trained on modern populations. This doesn’t necessarily mean it can’t be useful but when advertising it for anthropological uses I don’t think they’ve done enough there either.

Have they tried building a database removing Africans and anyone with African ancestry? Then tested it on Africans? That is, to simulate how well it works for populations they don’t have data on. It should be possible to find Africans and Eurasians or others with ancestry separated by at least 50000 (actually 100000) years. There’s a lot that can be done along these lines particularly where calculating ancestral distance would be useful.

I think that the sample size is quite small for anthropomorphic uses. It’s not just about the crude statistical criteria. There’s a lot of groups out there with genetic variations you want to tap. I can’t imagine sampling a thousand people would go far enough to provide a really powerful anthropological tool.

Regardless of what they could do better, unless I overlooked it I can’t see any of the kind of testing that you would want for anthropological studies nor the scale of samples that you would want from enough places (where sampling wants to be selective not just random).

The tool looks to me like a proof of concept at best even for forensic uses let alone anthropological. It could end up doing better for the latter but needs a lot more in the way of data, testing, SNPs and methods for predicting outcome.


What you’re seeing is nothing new. Scientific fraud has never actually been something we’ve been able to find a reliable way to put a stop to. It happens all the time, usually for sensationalism. Normal people getting a result out of the tool like blue eyes and very dark skin would immediately suspect it’s an anomaly, especially when the tool says 100% light hair but around 25% black hair.

Even if all scientists have good intent, fraud happens accidentally. Bizarre results tend to be more fetching in terms of the attention they receive and that they tend to be worthy of talking about. The default mindset should be that if your result is bizarre then there’s a good chance something is wrong.

This is one in a raft of combined political and commercial hoaxes. In this case, it’s an advertising scheme for I believe the National History Museum and a bunch of corrupt cronies. Basically a money making operation or a commercial operation. Yes, it happens in the press all the time. I’ve been part of that myself.

It’s also just a small part in a raft of lies and misconceptions. For example, that modern Europe, in terms of the modern borders represent historic Europe or the region. Misleading statements that we all came in from Asia.

Before this you had Kostenki, the “First European” which the media then made out was mixed race between Africans and Asians or something because why do we need to have a campaign to promote interracial sex, that’s a bit seedy isn’t it? I don’t know if anyone has looked at a top shelf recently but those efforts are redundant.

Then false history that sub-Saharan Africans have always been in Europe in large numbers because scientists can’t tell North Africa from sub-Saharan Africa.

It just goes on and on. If you see it in the media, it’s probably false. There’s no point relying on the newspaper because it belongs in the fiction section. It’s more reliably instead to just simply go back to word of mouth. Hearsay is more reliable than the media.


When you have a law enforcement tool that takes the DNA of a white person and turned them into a black suspect then I rest my case. If I committed a crime and law enforcement ran my DNA through this tool it would tell them they’re looking for a mixed race guy when in fact I’m fully white. I’m not happy with all this fake racism people are going on about when it comes to the police but a tool like this blackening up suspects is pretty damned bad. It’s a racist tool.

If the tool is based on prison inmates where there’s a large disparity then it might actually explain why it’s blackening them up. The problem is that scientists just can’t help being racist try hard as they might.

90% of what you read about in science is actually bogus. Real science tends to be boring or most people aren’t literate in the domain. Real science rarely makes the news. Fake science does.

For things like Cheddar Man, they’re never going to clone him to really find out, so they’re very unlikely to be proven wrong. Science is only reliable in cases where it can actually fundamentally be tested. For example, putting up rockets, where either they work or they don’t. Outside of things like that, science tends to quite often be pure trash.

Anyway, based on background data, that is the populations in Eurasia and their splits, we can’t really say exactly what Cheddar Man might of looked like. It’s not certain he was white but most probably close to it or not that far off Eurasian populations are today. As most people put it, a good chance of being olive coloured. Possibly similar to an Indian or Mediterranean or really any intermediate race.

Quite possibly, he had a unique tone and shade. Conversely, he may have been more white than people are today! We just can’t really say for certain. Him being dark black however (like the Andamanese) tends to be on the lower end of the probability scale. Europe has experienced changes in the environment particularly if you include things like North Africa as the full true European region. We can’t rule out entirely people going light then dark then light again.

Perhaps Cheddar Man was a natural born scientist like me. Did he have Asperger’s Syndrome? What would he have thought of all of this? Either way, you know a real scientist when although a rough cut, they actually publish a new scientific paper as an answer to a question on Quora.

I’ve updated this answer substantially as I’ve done my investigations. Some of it is anachronistic. Covering a topic then pursuing questions. I first tried this in when the story about Cheddar Man came out and the data. I vaguely remember that I didn’t find it very satisfactory because I could just change one allele and change the result. Many did that. It didn’t have a good consensus so just one misread or overlooked mutation or something and it’s a bust. I think I found the ASIP issue but didn’t have a baseline for comparison. This time around I actually had my own DNA on hand as a baseline.

Edited by DoTheScience
Link to comment
Share on other sites

image.png.30c9af634efac454663ddbff36d0adc6.png

 

image.png.31e288d1c76ef7c635cc07a3a6a886e4.png

 

So basically I'm assuming the university I caught out, and the NHM and the BBC uses their systematic racism to mass report me for telling the truth?

 

I scientifically PROVED, using MY OWN DNA that they were lying and it's taken down? This is cancel culture for you. I'm a white man whose DNA is just a little too politically incorrect because their "science" says I'm a mixed race man with blue eyes after paying through the roof to get sequenced myself. Ooops.

Edited by DoTheScience
Link to comment
Share on other sites

Due to this disagreement I have cancelled my sister who works at UCL and kicked her out of the family home. She is not welcome here anymore. She is cancelled. She's a stupid cat lady and she let her eggs dry up among other things and screams in my ear about peer review when what that really means is delete anything that proves you wrong. You cancel us, we cancel you. This is war. This is a war, a civil war, a family war and a national war. You are no longer family. You are not science, you are nothing. All they do is invoke the names of other people.

Edited by DoTheScience
Link to comment
Share on other sites

So the stupid cow tried to come back. Get the fuck out of our house. She's gone. You stupid fake scientific cow give up your PHD study and piss off you're not welcome in this house any more. You fake scientist anti family piece of scum bag liar. You're cancelled our of our family you stupid cow. You came back to make certain. Well let me make the message clear, you're gone cow. The UCL is your family now you lying criminal fake scientist cow.

Edited by DoTheScience
Link to comment
Share on other sites

Whine whine whine I didn't do anything, you cancelled me, I cancel you, whine whine whine all I did was being on the team i didn't do anything. Whine whine whine minorities and marginalised people. Shut up filth. I have other sisters you know. They're better than you are. You're a lost cause with your shrivelled ovaries.

 

You stupid cow, you lost your family. I hope it was worth it. Enjoy being a professor you stupid idiot.

Edited by DoTheScience
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...