Opinions about Nevgen Haplogroup Predictor

[Wing Genealogist]

https://www.nevgen.org/ has a relatively simple to use Y-SNP haplogroup predictor, where you plug in your Y-STR results and it predicts your Y-SNP haplogroup.

One tool I discovered is to click on the "Settings" (on the right side of the brown/grey box below the darker black bar at the top of the page). This will allow you to select certain subclades (R1b, R1a, I, J, E) as well as other options. Also, you can go to a FTDNA Project Y-DNA STR results page and copy the STR values and simply paste them into the white box below this brown/grey box.

This tool does give some information about the prediction, such as probability of the prediction being correct and some fitness scores. I have taken the Big Y-700 test and plugged in only 67 STRs and it came back with a Z6 clade prediction (I fall under a Z6 subclade). The probability was 96.56%

I have plugged in a variety of 67 & 111 STR results for kits that have done no SNP testing. The majority of the times, the SNP Haplogroup prediction has a probability between 95-100%. This seems to be rather optimistic to me, and I am wondering how others feel about the probability of predictions made by this tool.

[FionnSneachta]

I've never come across Nevgen being incorrect before. People usually get the result of an unsupported subclade if SNP testing hasn't been done their group. The prediction is for an SNP that is far back in time so it's not as if Nevgen can be used instead of Big Y but it will give someone an idea of where they should fit in the tree.

My dad gets a 100% prediction for R1b L21>DF13>Z39589>DF49>>ZP75. ZP75 is 7 SNPs above my dad's terminal SNP. All that I can say from my own experience is that their predictions don't seem to be too optimistic but seem to be rather accurate.

[Wing Genealogist]

If this tool is at all accurate (and I certainly hope it is) then it shows how much better haplogroup predictions can be than the EXTREMELY general predictions made by FTDNA.

[Paul333]

I put my 17 Str's in and it came up with ' unsupported or rare Haplogroup, at the top of the list was Y N1a1, with Snps L1026, Z4908, down to Y7795.

I think its OK for using the information from the common Haplogroups but once you have an uncommon one, with little information it seems to struggle, If maybe I had more STR results it might help, but having only a low number its not enough.