/sci/ - Science & Math

>>14489525
>>14489371
the lines are roughly this
asian
caucasian
african (potentially 2 types)
american
australian

no better definition of race/subspecies/breed exists than this
idk why people who advocate for race so much are this vague on this subject but typically they post this >>14491513 and go 'see race is real' without actually saying how many races and what they are
well those are the races as i've posted them and algorithmically classifying humans yields this as best estimate

>>12612968
proof btw

>>12606575
hmm rosenberg (2006) seems to put at k=5 east asians and native americans in separate groups with austronesian and papuan/australian int he same group. which is more accurate? i mean it couldn't be totally arbitrary... could it?

>>11681651
>No one said Fst's are the be all and end all of human sub calssification?
I've seen at is a talking point by many anons here, but also by Woodley a famous HBD blogger
Anyways, three's more than one way of showing it
>We could just go with taxonomy,
ok
>say these groups are physically separated by geography,
they're not, they overlap both geographically and genetically through logical locations such as asia minor, caucsus and the balkans
>exhibit noticeable differences,
not really
arab looking europeans are very common and so is the reverse - can't say the same for africans

>>11662825
why do you always post the absolute shittiest chart taken from a paper that isn't even about grouping but about black americans
>cape mixed
>hadza
>sub saharans
are those races we should accept universally or what
that paper's focus is completely different
also
>western eurasians/saharan africans
are those two synonyms or what

pic related is far more clear

>>11649701
>>11649697
ok then now that you've accepted clusters are valid way of doing races
>asian
>caucasian
>amerindian
>african
>australian

>>11236780
I posted that same plot in a different thread
You are essentially right, there are wrinkles here and there, but it is the closest you'll get to sub-species or races.
Something like
Caucasian, Asian, African, Australian, American

Also the lab can prove not just your race but if given good data to compare, it can probably prove which country or region you come from, which doesn't invalidate the concept of doing larger clusters.

>>11236394
>it's race denial to say arabs are white
lmao
It's not me saying it, it's literally every algorithm that clusters them with europe - it's a fact that the same ancestral populations gave rise to both

How about YOU tell me what the races are then

there seems to be a few polfags here trying to convince people middle easterners aren't white

middle easterners (and nafris) are most definitely the same race as europeans - genetics AND skulls both confirm this

>>11070032
based and racepilled
i have a similar chart

each bar is a person, i'm guessing it\s the same with your chart

different clusters as well

>we're all the same race

>>9381482
>Does human races/populations exist?
These studies need far bigger samples sizes to take into account regional continuity. Here's a kmean graph of genetic differences with different number of groups, it may be a little bit more serious even if it comes from Wikipedia
PCA analysis are prone to biases and you need really more and more diverse samples

>>9129166
https://www.newscientist.com/article/mg22129542-600-neanderthal-human-sex-bred-light-skins-and-infertility/

>Some of the genes, meanwhile, appear to have led to fertility problems. For instance, Sankararaman found that the X chromosome is almost devoid of Neanderthal DNA. This suggests that most Neanderthal DNA that wound up on the X chromosome made the bearer less fertile – a common occurrence when related but distinct species interbreed – and so it quickly disappeared from the human gene pool. “Neanderthal alleles were swept away,” says Sankararaman.

+ there are no neanderthal Y chromosomes in humans

basically 'thal-human people were less fit and possibly infertile and only those carrying harmless (straight hair?) or helpful(light skin?) alleles survived
>>9129172
>Why?
i think it's interesting to try to find the best clustering of humans, purely for algorithmic reasons

though it's pretty hard nowadays(post-colonial times) because half the world moved around and there is new, let's say, ethnogenesis happening in the americas
even before colonial times you have basically mixed populations in central asia, east africa ..

basically you'd need a % - let's say 70% or above ancestry and you are of race X and below 70% you are of mixed heritage?? i guess

it's pretty hard to do it

pic related is the best clustering (k=5), because it also happens to match continents
you'll notice that one of the bedouin sample is more african though lmao (each line is a person sampled if i remember correctly)

a sindhi dude is more than half african, while his neighbor is quadroon asian

but this is the best you can get in terms of races - 5 races, matching continents - caucasian, australian, native american, asian, african
at lower k's or higher k's you start getting weird shit

>>9075791
>races are like breeds guys xddd
Entirely false.
Make up your mind.
Either race matters or it doesn't matter.

Because when you ask a computer to organize populations into races you get pic related.

K=5 displays best discontinuity between populations AND it matches continental populations(as well as loosely matches earlier ideas about race when people used skulls).

It's insane the walls people build inside their heads to separate contradicting beliefs.
>Race matters
>Never mind computers clustering MENA with Europe

This is to prove to you that you guys need to abandon race, or move to the 'race is a social construct' wagon, because no good will come of 'race' - for your movement that is.

>>9019977
>rosenberg
this is precisely a report that clearly guarantees
k=5 reflects biggest jumps in continuity, but for 2,3,4,5,6 same result as far as this is concerned

an unbiased computer clusters euros and mena as THE SAME RACE
>>9019985
>All I see is different populations
Are you stupid?
look at figure 1

also you ignored my post
>if you control for age and class

>all these useless posters saying race exists but not even attempting to name the races
there are 5 continental races.
caucasian
african
east asian
american indian
australian

either that or there are 50 + races (which is the same as there are no races)

>>8829310
Not as retarded as some other racial classifications you can come across.

Of course I wouldn't call it subspecies, because we aren't that far separated(looking at Fst). But more importantly humans are sentient beings and have races. The term 'race' in modern days is usually used to describe humanoid, intelligent creatures - just like in sci fi. It's a more noble and more correct term than subspecies.

I prefer K=5, which sort of amounts to a similar thing, since human variation strongly follows how humans migrated (go figure).

>>8825944
good question

you take some SNP's, run a clustering algorithm(something like k-means) and you get this

I prefer k=5, since it matches continents, seems most meaningful