Which genetic company is the cheapest to take a test and is it worth? Or should I go for 23andme?

[Congolese Rice]

You understood me wrong. I mean, you can get to know your recent ancestry, even if you are a extremely mixed person, just by going back 3 or 4 generations in your family tree, and it will be much more accurate than any ancestry report out there.

Yes, while that might be true. You dont necessarily have to inherit the amount of DNA that you got from your ancestors, which is where autosomal dna tests come in quite handy. And also, if your family is from the caribbean like mine is, it tells little to nothing.

My great grandfather was born in Surinam, but he was a dutch. If i didnt know he was 100% dutch ethnically then i wouldve been thinking all types of shit, like for example that he was black (slave trade) asian, or native american.


i think it heavily depends upon the individual, but personally i dont think autosomal dna tests are working out too well for me, then again thats just my opinion.

[Brás Garcia de Mascarenhas]

Your first sentence is completely wrong, the more SNPs the more accurate your ancestry prediction will be, i can enter in details if needed.

Your raw data will not change much from test to test, because they test similar numbers of SNPs. And 'no haplogroups' is incorrect, you can get your Y-DNA using some basic prediction tools because MyHeritage do test some haplotype snps.

Can you name me an ancestry tool that reads more than 25%-30% of the total number of SNPs provided in your raw data? A tool that is capable of reading more SNPs can give you more accurate projections but it doesn't mean that all those 700,000 SNPs are relevant for ancestry because they are not. Your Raw data shouldn't change indeed much from company to company, they all use the Illumina chip with subtle differences.

You understood me wrong. I mean, you can get to know your recent ancestry, even if you are a extremely mixed person, just by going back 3 or 4 generations in your family tree, and it will be much more accurate than any ancestry report out there.

Yes but Bell Beaker is a native Portuguese, he can certainly go 3 to 4 generations back and its all native Portuguese as well.

[Profileid]

Do either 23andme or ancestry. Nothing else is worth spending money on.

[Brás Garcia de Mascarenhas]

What about FamilyTreeDNA? I mean, I was satisfied with the results.

23andMe and FTDNA have the best and most comprehensive platforms in my opinion, with recurrent updates (FTDNA even allows you to go deeper when it comes to haplogroups and to join research projects). When I pay 100$ I want more just than raw data, but that's just me. Genetics is a field that is ever growing and ever changing and I expect these platforms to improve with time and to refine their data and results.

[Token]

Can you name me an ancestry tool that reads more than 25%-30% of the total number of SNPs provided in your raw data? A tool that is capable of reading more SNPs can give you more accurate projections but it doesn't mean that all those 700,000 SNPs are relevant for ancestry because they are not. Your Raw data shouldn't change indeed much from company to company, they all use the Illumina chip with subtle differences.

Do a test: delete a random 200.000 SNPs from your raw data and upload it to DNAland, you will see how much your ancestry report will change. With less markers tested, more imputation of the 'gaps' in your genome is needed, and less accurate will be your results. I wonder where you got this '25-30%' from, almost all markers are relevant for an accurate report.

Yes but Bell Beaker is a native Portuguese, he can certainly go 3 to 4 generations back and its all native Portuguese as well.

If he is not interested on deep ancestry, then i don't see the motive for a genetic test. They will tell him what he already knows, that he is 100% Portuguese (considering that he have no foreign ancestry in the last 3-4 generations, which is always a possibility).

[Brás Garcia de Mascarenhas]

Do a test: delete a random 200.000 SNPs from your raw data and upload it to DNAland, you will see how much your ancestry report will change. With less markers tested, more imputation of the 'gaps' in your genome is needed, and less accurate will be your results. I wonder from where you got this 30% from, almost all markers are relevant for an accurate report.


If he is not interested on deep ancestry, then i don't see the motive for a genetic test. They will tell him what he already knows, that he is 100% Portuguese (considering that he have no foreign ancestry in the last 3-4 generations, which is always a possibility).

Check how many SNPs most Gedmatch tool\projects read on your Raw Data, it is never even slightly close to the total number of SNPs provided on your Raw Data, no matter the chip.

Things are not so simple and linear as you say though. Take me and my girlfriend for example, our family villages are separated by only 20 km, we both are native Portuguese and so were our three to four last generations yet our genetic makeup is different.

Having foreign ancestry while being from the country side in central Portugal like me and Bell Beaker is extremely unlikely considering that the villages were so remote and most people lived all their life in a 10 miles radius and didn't knew anything else. You don't even have to go 3-4 generations backwards to understand that, my grand-grand-mother was born in 1904 and she only saw the Ocean for the first time in the 80's or something (which for Brazilian standards is just a step away, 150 km). Portugal since the Reconquista had a few invasions and wars but nothing comparable to the majority of other regions in Europe that had by far more turbulent and unstable historical periods in recent history.

[Iloko]

23andme ftw!

[Brás Garcia de Mascarenhas]

Do a test: delete a random 200.000 SNPs from your raw data

Most amateur projects available to common mortals like us don't even read 200.000 SNPs to begin with, you are lucky if they reach 160.000 SNPS so obviously if you delete that many it will have a big impact. What I mean is that from 700.000 SNPs not all of them are necessarily relevant or have an impact on the comprehension of your genetic makeup.

It is a different story though if you want to go in depth on health reports or upload your data to platforms like Promethease.

[Ruderico]

I got 23andme for the gf and I, and we're pretty happy with the results. It's fun to see stuff with GEDmatch and Global25 aswell, and you get then get these cheaply. Eventhough 23andme isn't the best for GEDmatch, it still gets the job well done if you convert to v3. I got rather consistent results in pretty much all models

[Hyoga7]

Boas,

I have results from the Big 4: National Geographic; FTDNA; 23andMe; My Heritage.

Haplogroups: the best was National Geographic (it went deeper); FTDNA and 23andme started with R L-21 and a couple of months later changed to different/"smaller" subclades.

Autossomal: National Geographic seemed very ancient; 23andMe seemed very recent; FTDNA and MyHeritage were the ones that were less European.
I dont't wanna go into conspiracy theories, but I have read somewhere that MyHeritage was Israeli and thus more interested in showing results as less European to prove that Europeans aren't 100% Europeans. Truth or Lie?

I totally understand the $ factor. Try to take advantage of promotions around specific dates (Christmas, Black Friday). Do you have family or friends in the USA or UK (it's generally cheaper in these 2 countries) that can bring you the test from there?


Boa sorte!