This calculator can be interpreted in many different ways if you take out some samples and/or re-label some samples, but it works in a very interesting way.
Some components can be modeled by others in a very good way:
It puts me too much into Anatolian and too little into Levantine; thus, I get Arabic and Italic to balance it out, despite them not making much sense. But also it is a European-focused calculator & meant to be used with ADC for a reason xd
It puts me too much into Anatolian and too little into Levantine; thus, I get Arabic and Italic to balance it out, despite them not making much sense. But also it is a European-focused calculator & meant to be used with ADC for a reason xd
Yeah, both me and my father get weird distances with G25 calculators & samples, my closest modern population average is something like 3.9. My friends of a similar but slightly different background get normal fits, and we all took 23andMe at the same time. I think it is because my area is heavily Anatolian/ANF whilst being Levantine and showing up this way on commercial tests— I'm about the same distance to Greek Cypriot as I am to Lebanese Christian. On this set PCA'd it kind of ends up like this:
Yeah, both me and my father get weird distances with G25 calculators & samples, my closest modern population average is something like 3.9. My friends of a similar but slightly different background get normal fits, and we all took 23andMe at the same time. I think it is because my area is heavily Anatolian/ANF whilst being Levantine and showing up this way on commercial tests— I'm about the same distance to Greek Cypriot as I am to Lebanese Christian. On this set PCA'd it kind of ends up like this:
from my experience 23andMe data made my distances quite far too. I believe mine went from .0256 with ancestrydna data to about .033 with 23andMe. It is a bit of a big jump.
from my experience 23andMe data made my distances quite far too. I believe mine went from .0256 with ancestrydna data to about .033 with 23andMe. It is a bit of a big jump.
I'm going to do an Ancestry test soon and hopefully this will fix some issues. I think it is a bit odd that it was both me and my father (I actually tested him later than I did with the other people) that had this problem while people with the same 'ethnicity estimate' did not, but regardless I am interested in having higher quality raw data, so I can at least eliminate that as an issue.
Yeah, both me and my father get weird distances with G25 calculators & samples, my closest modern population average is something like 3.9. My friends of a similar but slightly different background get normal fits, and we all took 23andMe at the same time. I think it is because my area is heavily Anatolian/ANF whilst being Levantine and showing up this way on commercial tests— I'm about the same distance to Greek Cypriot as I am to Lebanese Christian. On this set PCA'd it kind of ends up like this:
Interesting, u seens kinda Cyprus shifted for a Levantine
I am having fun with this and bumping the thread because it feel, so I will keep going a bit haha.
Spoiler!
Seems like it might just be a population quirk, considering my Dad has some element of it too, though he is slightly more aligned with the other Levant samples. Honestly, what I have gotten out of this is that I probably should have tested my mom as well... https://i.ibb.co/NmcRmRF/sacha-5.jpg
(Sacha_LF2 has 1.2% Cypriot on 23andMe but that doesn't mean much, thought it was interesting to note though.)
On Eurogenes K36 and K13 I have very high East_Med. I think it might be high ANF/EEF that does this. I will probably mess around with qpAdm a bit soon to get a better reading on that, but on top of G25 component calculators giving me heavy ANF, I also seem to get small Italian splits that are almost always Sardinian. And then when you look at 2D-ified PCAs, I leave the main population 'packs' in a weird way, I have posted this PCA a few times before, but it is relevant and I like the visual:
But if you see on that global one where I have zoomed in on the Europe and West Asia clusters, it kind of looks like I could be in that overlap zone with other Near East populations and South Europeans, but something is pulling me 'out.' And that other population in the bottom left corner is a little cluster of Sardinians, who have very high EEF. I don't care as much about the G25 calculators as a measure of admixture, they are not the most accurate in that regard, but as a comparative tool for shift they are pretty convenient.
I also wonder if back-to-Syria Maronite migration from Cyprus had any effect on the population in my area specifically, because there are many with a last name that means "Cypriot," but as far as I can tell those who returned didn't really mix that much outside of the Maronite community. I am related to a few on 23 that show up as 100% Levantine and have multiple generations tested. But I guess that is just the mystery of Sacha Fit Syndrome (trademark)
01-11-2024, 08:01 AM
Solitude
Jinglebell your coords raw data is from what company? Myheritage?
