PDA

View Full Version : WeGene accuracy...



NiamhWitch
06-06-2018, 11:24 AM
I recently uploaded my 23andMe results to WeGene and got some rather crazy results. The ancestry was all... well, wrong. Just straight up wrong. However, I was excited that it provided more specific haplogroup predictions than 23andMe did. It narrowed down my mtDNA from H3 (what 23andMe said) to H3v2. Any way I can check the accuracy of this?

Now, I also uploaded my father's 23andMe results, and was excited that they also provided a more detailed Y-hap for him. 23andMe assigned him I-M253, whereas WeGene narrowed that down to I-Z2535. That said, I went through my dad's raw files and went through the corresponding SNPs for I-Z2535 and others. His raw files show that he is NEGATIVE for I-Z2535 (though he is positive for its parent, I-Z140). How on earth did WeGene determine he was Z2535, when he definitely isn't? It also makes me doubt their prediction of my mtDNA.

On the one hand, I was still able to definitively narrow down the Y-haplogroup to I-Z140, which I wouldn't have known to do if I hadn't uploaded to WeGene, so I'm grateful for that... but on the other hand, they gave me all sorts of wrong information, so that's kind of annoying...

nuplix
06-06-2018, 11:28 AM
I recently uploaded my 23andMe results to WeGene and got some rather crazy results. The ancestry was all... well, wrong. Just straight up wrong. However, I was excited that it provided more specific haplogroup predictions than 23andMe did. It narrowed down my mtDNA from H3 (what 23andMe said) to H3v2. Any way I can check the accuracy of this?

Now, I also uploaded my father's 23andMe results, and was excited that they also provided a more detailed Y-hap for him. 23andMe assigned him I-M253, whereas WeGene narrowed that down to I-Z2535. That said, I went through my dad's raw files and went through the corresponding SNPs for I-Z2535 and others. His raw files show that he is NEGATIVE for I-Z2535 (though he is positive for its parent, I-Z140). How on earth did WeGene determine he was Z2535, when he definitely isn't? It also makes me doubt their prediction of my mtDNA.

On the one hand, I was still able to definitively narrow down the Y-haplogroup to I-Z140, which I wouldn't have known to do if I hadn't uploaded to WeGene, so I'm grateful for that... but on the other hand, they gave me all sorts of wrong information, so that's kind of annoying...

Don't rely on WeGene for ancestry breakdown. GEDMatch and GenePlaza are far better options.

NiamhWitch
06-06-2018, 11:36 AM
Don't rely on WeGene for ancestry breakdown. GEDMatch and GenePlaza are far better options.

I'm not relying on them for the ancestry. It was so far off, lol.
My question was really about how they could get the haplogroup wrong, ya know? :confused: Especially when I can go through the SNPs manually one-by-one and prove it. It just seems odd that they would get that wrong. I'll have to ask someone in the mtDNA boards how I can verify my hap the same way I did for my father's...

I already have our results uploaded to GEDmatch (Genesis, because of the V5 chip...) :)

deadly77
07-13-2018, 08:28 AM
Regarding the mtDNA, you can apply the same approach that was used for the Y DNA - look up the individual SNPs in the raw data. Look up the mtDNA phylotree here http://phylotree.org/index.htm - then look up the defining SNPs that lead from H3 to H3v, H3v1, H3v2. Look up the #rsid for those SNPs, then go and look in your raw data.

You can also do this more quickly by running your raw mtDNA FASTA file through the mthap site by James Lick which can be found here: https://dna.jameslick.com/mthap/ - this site is free, easy to use and does not store your data. It'll tell you how much of your mtDNA was analyzed by 23andme (in my case 19.7%, so not so much), will colour code your SNPs so you can see no calls, mismatches, extras in addition to matches for the proposed haplogroup. It also uses Build 17, the latest version of the mt phylotree - I believe 23andme uses an earlier build. No idea what build WeGene uses.

gjenetiks
09-17-2018, 03:19 AM
Wegene was also very inaccurate for me. I got a quarter spanish.

gjenetiks
09-17-2018, 03:29 AM
Isn't Wegene a test more geared to Asians?

jshook
09-20-2018, 08:43 PM
Unless you're Asian, Wegene is fairly useless for you from an Ancestry standpoint and, at least in my case, the the maternal haplogroups for H folks are terrible. I have a Japanese grandparent and it called the Asian parts relatively correctly with a little weirdness, such as some unknown Chinese ancestry (screenshot attached). My Euro ancestry was hilariously terrible.

As far as haplogroup predictions went, both my Ancestry and 23andMe uploads it correctly predicted R1b1a1a2a1a1c2b2 or R-Z9 for my paternal subclade (which matches both 23andMe and LivingDNA's results.) But for my mom's side, the Ancestry upload stopped at R0 and the 23andMe result was just R. I'm H3.

26080

czl
09-21-2018, 01:16 AM
I like wegene for the calculators it has, not the actual results, and at least I got hapolgroups although they weren’t very precise. Also as a test that’s Asian centric, they should really get some more groups, my Kazakh dna wasn’t detected because of the lack of the region, however they did get my Uzbek.

mildlycurly
10-12-2018, 07:30 PM
Wegene is aimed at the Asian market, so expect it to be quite inaccurate for non-Asian admix right now. I do have some minor South Asian admix so I will take a Wegene test as soon as I can get my Ancestry file to upload!!!!!!! >:(, but I'm not expecting accuracy with regards to my European results.