SO MY ANCESTRY DNA RESULTS WERE WRONG? | Ancestry DNA Update (Afro-Caribbean)

[@crescendyr8438]

Ethnicity tests are not objective. They take samples from various regions and whatever segments of DNA are found the *most* in a particular country, that segment of DNA is labeled as being of that country. That doesn't mean that same DNA segment is not prevalent in other neighboring countries at a lesser frequency. So if in 2019 they found certain DNA segment at the highest frequency in Benin your results will show that. But in the following year, after testing more people in that region and increasing their sample database size, they found that same DNA segment is actually more prevalent in Nigeria. So now the *arbitrary* label they assigned that segment switches from Benin to Nigeria. The DNA is not changing; just the labeling. Keep in mind that these countries are just arbitrary lines on a map determined by European colonizers. In reality Benin, Nigeria and Cameroon, for example, are all the same people. The truth is the results are more accurate if they just said West Central Africa, but that doesn't tell people what they couldn't already guess and people will complain because it's not specific enough. So they give you these results that are just educated guesses that will keep changing every year because the detailed country by country breakdown looks _"sexier"._

[@LaurelJade]

Hey, thank you so much for your insightful comment. Yes, I totally agree with you. I guess ethnicity tests aren’t for those looking to specific and accurate results. In my case, I just know I’m linked to West Africa - which is obviously no surprise. Thank you for watching. Blessings 💕

[@anmariee5560]

I guess I just wasted my money on 23 and me then.

[@crescendyr8438]

​@@LaurelJade What I have found regarding AncestryDNA is that the DNA Relatives make up the most important aspect of the test. Because those are actual relatives with DNA that matches your own. They help to put a people to the regions. I have combed thru my DNA relatives and found 100% African (or half-African/half something I don't have in common) that tell me what ethnicities I stem from. I have found 16 matches so far, not including Half-African/half-Diaspora because I don't know which half I'm related too. Of the ones 100% African or those I know I'm related to the native African side I found 1 Wolof of the Lion Clan from Dakar, Senegal. His people are where my "Senegal" comes from. 1 Temne from Sierra Leone, which is where my "Mali" comes from. 1 Hausa from Kano State in northern Nigeria where some of my Nigerian comes from. 4 Ewe from Ghana and Togo, where my IvoryCoast/Ghana and Benin/Togo come from. And finally 9 Igbo mostly from Anambra State in SE Nigeria where I'm comfortable saying most of my Nigerian comes from. That covers all of my West African results. I'm still hoping to find someone from the southern half of Africa to account for the remaining Bantu portion. I really wish Ancestry had a chromosome painter like 23andMe so I could see exactly where the relationship is from. As I had a guy who was only 30% Nigerian an 70% European. But because 23ndMe had the chromosome comparison tool I could see that I was related thru the Igbo grandfather (the 30%) and not the European. If that guy took AncestryDNA there would be no way to know. AncestryDNA has *WAY* more DNA relatives than 23andME. I get more relatives in one day from AncestryDNA than I get on 23andMe in a whole month.

[@crescendyr8438]

@@anmariee5560 See reply above to Laurel Jade. These tests aren't useless. If you want to get a breakdown of what regions you come from they will give you that. If you want something more in depth it takes some work, but you can find that too via DNA Relatives.

[@anmariee5560]

@@crescendyr8438 Ok then. I'll look into that next time.

[@stefanorossi9643]

Benin borders with Nigeria, and in Nigeria there is a city called Benin City. The people of those two countries may have common origins. A long time ago, the African national borders were different because there were strong African kingdoms.

[@shryve]

As more Africans and Afro people in the Americas get tested, the data gets more accurate.

[@LaurelJade]

They do say that. There is another update, I’ll be making a video on that real soon! 😅

[@drgbengaa.shadare7809]

Laurel Your DNA has not changed. Remember that the peoples of Nigeria, Benin, Togo especially (to some extent Cameroon) are similar. Actually Benin and Nigeria and Togo could have easily been one country if not for colonisation.

[@LaurelJade]

Hey Dr Gbenga, Thank you for watching and commenting. I understand that my DNA hasn't 'changed' per se. I know that Benin, Nigeria, Togo all border one another. However, in my opinion since my DNA results have changed, Ancestry DNA does not give accurate results.

[@curtiswilliams8285]

@@LaurelJade they "changed" because they got more sample populations for the regions. DNA testing is still a new thing so as they get more sample populations, they'll be more updates. As the poster said, your DNA is still your DNA, that won't change. It's the greater understanding of interpreting it and that's going to be continual

[@the_ohioplaya]

PCA is a mathematical transformation that reduces the dimensionality of the data to a smaller set of uncorre- lated dimensions called principal components (PCs), which has numerous applications in science. In population genetics alone, PCA usage is ubiquitous, with dozen standard applications. PCA is typically the first and primary analysis, and its outcomes determine the study design. That PCA is completely non-parametric is the source of its strength. Any genotype dataset can be rapidly processed with no concerns about parameters or data validity. It is also a weakness because the answer is unique and depends on the particular dataset, which is when reliability, robustness, and reproducibility become a concern. The implicit expectation employed by PCA users is that the variance explained along the first two PCs provides a reasonable representation of the complete dataset. When this variance is minuscule (as often with human populations), it poorly represents the data. Rather than consider using alternative analyses, authors often choose not to report the variation explained by PCA. Regardless, it is not a proxy for the reliability of the results. Here, we carried out extensive analyses on twelve PCA applicaitons, using model- and real-populations to evaluate the reliability, robustness, and reproducibility of PCA. We found that PCA failed in all criteria and showed how easily it could generate erroneous, contradictory, and absurd results. This is not surprising because PCA is blind to the data and their meaning. The covariance matrix is calculated from the centered matrix itself created simply by subtracting the mean A, from the original matrix A, disregarding the weights and geography. The remaining transformation consists of the dimensionality reduction, which is less problematic; however, that the first two PCs that capture most, but still a very small part of the genetic variation, are typically analyzed creates further misinterpretations. Given the omnipresence of PCA in science, an intriguing question is whether multidisciplinary PCA results should be reevaluated? Based on our analyses and critical evaluations published elsewhere, we cannot dismiss this possibility. As PCA lacks any measurable significance or accuracy, we argue that its dominance in population genetics could not have been achieved without the adoption of two fallacies: cherry-picking or circular reasoning (i.e.. "exploration"), the screening and selecting PCA scatterplots that fit preconceived hypotheses while ignoring the other plots, and the a priori where PCA results are interpreted based on pre-existing knowledge because PCA scatterplots are uninformative a posteriori. As a "black box" basking in bioinformatic glory free from any enforceable proper usage rules, PCA misappropriations, demonstrated here for the first time, are nearly impos sible to spot. The fact that population affinities vary appreciably between closely related, ostensively equivalent datasets is deeply worrying (PCA applications were cited 32,000-216,000 times). Researchers from adjacent fields like animal and plant or medical genetics may be even less aware of the inherent biases in PCA and the variety of nonsensical results that it can generate. We consider PCA scatterplots analogous to Rorschach plots. We find PCA unsuitable for population genetic investigations and recommend reevaluating all PCA-based studies.

[@westindianmalkah8570]

Benin/togo can include the yoruba side of Nigeria southwest nigeria, and the togo can include parts of ghana particularly the ewe people of the volta region. Your dna is the same many are bordering each other and consist of same people pre colonial times. On ancestry look at the maps it will show all included in your percentages

[@LaurelJade]

Heyyy, thank you for watching! I totally get that the West African countries that make up my percentages border one another and are very close. However, what I'm saying is that the percentages are not precise - as they keep changing with my highest percentage being Ghana, Nigeria, Benin and Togo etc. So all I'm saying is, if someones looking for just one country that they originate from, Ancestry DNA is not very helpful in assisting with that. Blessings 💕

[@westindianmalkah8570]

@@LaurelJade I totally agree with you. I had 4 different readings from ancestry and to me the most accurate was the very first one. I don't really like their ethnicity readings I prefer 23 and me 😊

[@selasedu]

Benin Togo means Ewe and Fon people

[@blackmagic6]

As I've seen already written in your comment section, there is a high likelihood that your overriding tribe is Yoruba. I say this because the Yoruba tribe is the tribe that historically and still exists in the region that is South West Nigeria and Benin and Togo. Another reason is that I'm also Yoruba who also did a DNA test out of curiosity. My results came back 81% Nigerian and 19% Benin and Togo. When I informed my parents they were surprised as I was about my result. We attribute the result to the break up of the Yoruba tribe after colonialism. Finally, your results are consistent with many results that I've seen from people of Jamaican descent. I think that the prevailing notion that most Jamaicans originate from Ghana needs to be revised because the DNA results do not bear this notion out. Peace.

[@LaurelJade]

Wow! That's so interesting. I know quite a bit about the Yoruba tribe already. It's true though that many Jamaican's are linked with Ghana and the Akan tribe. But I'm sure this isn't the case for everyone evidently. Funnily enough my DNA results have changed again, I'm going to do an updated video soon. Ivory Coast/Ghana has gone back up to being a high percentage. Thank you so much for watching! ☺️

[@blackmagic6]

@@LaurelJade: Thanks for your reply. If you remember, could you send me a link when you do your update? I'm curious to see what your new results are. Peace.

[@moimoimoi2585]

Nigeria and Benin are the same ethnic groups ( Yoruba) Togo and Benin have the Fon people Ghana and Togo have Ewe people Ivory Coast and Ghana have both Akan people Cameroon Gabon Congo And more have same ethnic groups too AFRICA country borders was create by the european. AFRICA is about ethnic groups and tribes. You can find in differet country the same ethnic groups. So DNA IS RIGHT

[@LaurelJade]

It's all very interesting when you put it like that. Thank you!

[@RAS_DNA]

A ethnicity estimate results should be taken regionally and on a continental level, not present day nation borders which in many cases have not existed for 100 years.

[@LaurelJade]

I agree! 💯

[@stayalert7060]

Take 23and me. I'm really interested at the Mali percentage cus that can easily link into Senegambia and also Sierra Leone.

[@LaurelJade]

Hey, yesss I’ve definitely been convinced to do 23andMe. Thank you 💕

[@kendal4452]

What is the difference between Ancestry and 23 and me?

[@patrickwright6644]

I love your hair style

[@LaurelJade]

Thank you :)

[@sr2291]

DNA tests are just estimates. Just because your family lived somewhere does not mean their DNA is from that area. Ancestry got rid of my North African results and are not reporting my small amount of Indian (India) DNA.

[@LaurelJade]

Hey! Yep, I totally understand that it’s just estimates and I’ve definitely learned a lot from doing this Ancestry DNA test.😌

[@sr2291]

@@LaurelJade Me too.

[@layahYasharahla]

Great updates it’s cool my test results are about 3 weeks old I doubt they change but who knows lol 😂

[@LaurelJade]

Heyy, it will be so interesting for you to see... trust me! Give it about 6 months LOL, come back and let me know too 😅 Thank you for watching💕

[@layahYasharahla]

@@LaurelJade lol 😂 I do KZbin as well feel free to check out my results about four videos down and I will lol I subbed to you also

[@dopevernacular2925]

Benin & Nigeria are connected, so it makes sense that those 2 would go up and down as more data is available.

[@LaurelJade]

Yep! It’s just strange because at first I didn’t have any Nigerian as a percentage and then it became so high lol.

[@DieezahArts]

AncestryDNA is not the company you want in terms of accuracy. The problem with them is 1/they don't seem to have enough samples from people like us to begin with so even if the science they're using is "normal" compared to other companies in that industry, the estimates they're going to give you will be more vague than a company like 23&me. 2/they only test your 22 first pairs of chromosomes unlike 23 and me, which also makes results less precise in terms of the information you can get from getting your DNA tested. I've used 23&me to test my DNA and later uploaded my raw DNA data to Ancestry to compare how they analyzed the DNA data... Not precise at all even if it went more or less in the same general direction as my 23&me results. The analysis of the 23rd pair gives you info on your patrilineal heritage if you're a guy or if a father or brother also gets tested. Ancestry is only there trying to make everything a cash grab. Now you can still upload your raw DNA data to Ancestry from other sources but they no longer give you the ethnicity estimate after you do as they used to, unless you pay (meanwhile they use your uploaded data to expand their database of samples used to lure new customers to them...) 🤷🏾‍♀️ They have a less diverse customer base. You're likely to find more matches on 23...

[@LaurelJade]

Woah, thank you very much for your comment, very insightful and makes me want to try 23&me. I agree with you, Ancestry DNA doesn’t seem to have enough samples from our people, which explains all the ever-changing updates. It would definitely be interesting to compare my results from both Ancestry DNA and 23&me. Thank you for watching! 💕

[@dominiquehudson8077]

Ancestry is definitely accurate in testing DNA. Both Ancestry and 23andme are just estimating, which means there's always a margin of error and possible updates. Neither company can truly get country-level labels right since most regional areas tend to descend from one major branch group. On top of testing 22 chromosomes vs 23, this deals with the X and Y chromosomes being more stubborn than the others. The 23rd pair tends to span back multiple generations; however, can literally be changed by the introduction of a single individual. This can make the 23rd pair highly inaccurate (although 23's matching system is more restrictive, which balances it out), which is why not every company analyzes the 23 pair. Definitely not something to worry about if you're just looking at admixture breakdowns. Regional percentages will stay very much similar + ancestry has a larger database with more tools for searching records. When it comes down to matching Ancestry is much better, but 23andMe is better at traits and health. P.S. You can't upload 23andMe data to Ancestry. That's never been an option. You're thinking of a different site. Ancestry has never accepted 3rd party uploads of raw data.

[@Elias_Truth]

@@dominiquehudson8077 i think ancestry is better. 23andme base percentages off of relative finder

[@richrosegold610]

@@dominiquehudson8077 I doubt it. Why is it that almost every black person I've seen on here with Ancestry results have almost the EXACT SAME RESULTS? They keep using the same 10 countries in an African Continent of 54 countries for results for everyone that is black. Every black person I have seen has near the same results as me. My father and his entire family is INDIAN and I mean Northern and Southern Indian, yet they only put me at 1% Northern Indian. My mother is born in the Caribbean, like me, and we have Arawak Indian blood, along with British and Irish ancestors, going back to the old records from the 1700s, as far as I know. They put me at 32% Nigerian and 44% of 7 other African countries, and I'm just like "WTF". This just doesn't sound right at all. Too many people have complained about it, and honestly I think it's bullshit. Apparently 23 and Me is much better, so I will be buying a kit from them soon.

[@dominiquehudson8077]

@@Elias_Truth 23andme uses allele mapping/clustering and datasets not the relative finder for their precentages.

[@philipbutler6608]

You have DNA from millions of people. No ethnicity for DNA is 100 accurate. Remember people have been migrating within Africa since your Ancestors were brought to the Americas. They don’t know who lived where in Africa 400 years ago. I think the neatest part of AncestryDNA is doing your family tree. It makes finding record’s easier.

[@LaurelJade]

I'd definitely one day want to make a family tree. Thank you so much!

[@philipbutler6608]

@@LaurelJade you know the 1940 census will help you a lot. It’s important to fill out the birth date and places and the death date and places. Don’t feel alone that you can’t find records DNA matches will help you with that. My grandfather was put out for adoption because he was born out of wedlock. A great grandmother was orphaned in the Civil War can’t figure out her parents yet either. And I know my grandfathers father but his mothers family is a mystery. So it’s just connecting the dots with DNA matches.

[@the_ohioplaya]

PCA is a mathematical transformation that reduces the dimensionality of the data to a smaller set of uncorre- lated dimensions called principal components (PCs), which has numerous applications in science. In population genetics alone, PCA usage is ubiquitous, with dozen standard applications. PCA is typically the first and primary analysis, and its outcomes determine the study design. That PCA is completely non-parametric is the source of its strength. Any genotype dataset can be rapidly processed with no concerns about parameters or data validity. It is also a weakness because the answer is unique and depends on the particular dataset, which is when reliability, robustness, and reproducibility become a concern. The implicit expectation employed by PCA users is that the variance explained along the first two PCs provides a reasonable representation of the complete dataset. When this variance is minuscule (as often with human populations), it poorly represents the data. Rather than consider using alternative analyses, authors often choose not to report the variation explained by PCA. Regardless, it is not a proxy for the reliability of the results. Here, we carried out extensive analyses on twelve PCA applicaitons, using model- and real-populations to evaluate the reliability, robustness, and reproducibility of PCA. We found that PCA failed in all criteria and showed how easily it could generate erroneous, contradictory, and absurd results. This is not surprising because PCA is blind to the data and their meaning. The covariance matrix is calculated from the centered matrix itself created simply by subtracting the mean A, from the original matrix A, disregarding the weights and geography. The remaining transformation consists of the dimensionality reduction, which is less problematic; however, that the first two PCs that capture most, but still a very small part of the genetic variation, are typically analyzed creates further misinterpretations. Given the omnipresence of PCA in science, an intriguing question is whether multidisciplinary PCA results should be reevaluated? Based on our analyses and critical evaluations published elsewhere, we cannot dismiss this possibility. As PCA lacks any measurable significance or accuracy, we argue that its dominance in population genetics could not have been achieved without the adoption of two fallacies: cherry-picking or circular reasoning (i.e.. "exploration"), the screening and selecting PCA scatterplots that fit preconceived hypotheses while ignoring the other plots, and the a priori where PCA results are interpreted based on pre-existing knowledge because PCA scatterplots are uninformative a posteriori. As a "black box" basking in bioinformatic glory free from any enforceable proper usage rules, PCA misappropriations, demonstrated here for the first time, are nearly impos sible to spot. The fact that population affinities vary appreciably between closely related, ostensively equivalent datasets is deeply worrying (PCA applications were cited 32,000-216,000 times). Researchers from adjacent fields like animal and plant or medical genetics may be even less aware of the inherent biases in PCA and the variety of nonsensical results that it can generate. We consider PCA scatterplots analogous to Rorschach plots. We find PCA unsuitable for population genetic investigations and recommend reevaluating all PCA-based studies.

[@nicolejackson2717]

Yes mines changed as well which I was very questionable about it 🧐

[@LaurelJade]

Get ready for it to change a few times, at least once a year lol

[@deeallen1526]

Just curious, where were you born and raised?

[@LaurelJade]

Heyy, I was born and raised in London, United Kingdom 🇬🇧😊

[@John-ji9tq]

It’s safe to say you are Benin and Togo

[@LaurelJade]

Really? 😅 Lol, well I’m not too sure when it comes to Ancestry DNA.

[@John-ji9tq]

@@LaurelJade it’s accurate

[@LaurelJade]

@@John-ji9tq if ancestry call it 'estimates', I'm sure that it can't be 100% accurate though?

[@landmark22]

Wait that's a major change

[@LaurelJade]

I know right! And there’s been even more of a change again!!

[@landmark22]

@@LaurelJade I did African Ancestry aswell

[@alexusbrooke]

U should try afroroots dna

[@LaurelJade]

Oooh, I'll check it out. Thank you! 💕

[@bluejay9968]

Ancestrydna is not good.

[@LaurelJade]

Oh dear, why do you say that? What other DNA testing do you think is better?

[@bluejay9968]

@@LaurelJade 23andme!

[@stayalert7060]

@@bluejay9968 Agreed 23andme is good. For example I'm Congolese half mukongo from DRC 🇨🇩 and it picked it up immediately 50% Angolan and Congolese and European for me is Portuguese. I've seen Haitians and Dominicans taking 23andme and finally their BaKongo DNA is showing and reflecting their culture. I often feel like Ancestry DNA forgets central Africans. All Bantu tribes are not the same genetically. BaKongos especially (coastal) are genetically closer to West African populations than other bantus in DRC or Angola. Ancestry is useless because it just lumps people together. Even Benin/Togo.. A Yoruba can get that as they are lingustically and culturally closer to tribes in Benin than to other Nigerians like Igbos or Hausas. 23 and me keeps updating which is good for us as central Africans especially BaKongos. We've seen Haitians 40% Angolan and Congolese which is accurate especially as their side of Haiti is 100% more BaKongo 🇦🇴🇨🇬🇨🇩🇬🇦influenced!

[@kendal4452]

@@stayalert7060 What is the difference between 23 and me and Ancestry? I'm new to both.