Batch prediction service of NevGen Predictor

Since September 2021, NevGen Predictor offers commercial service for batch prediction of haplotypes.

If you have hundreds of haplotypes (up to three thousands) which you need to predict, and you do not have time to manually enter them into our Predictor, you can send us haplotypes, and we shall do for you batch prediction on our computer, for small fee. That can save much of your time, at no big cost.

All haplotypes are first predicted through General Level of NevGen Predictor. Then, depending of prediction, if it points to any haplogroup that has its own specialized Level of prediction (haplogroups E, G, I, J, N, R1a, R1b) then it is predicted in the haplogroup-specific Level too. This way, most of haplotypes will be predicted by two Levels of Predictor, the first General and second which is haplogroup-specific.

Lets see example with only three haplotypes:


15 23 15 10 15-15 11 13 11 14 12 32 16 8-10 11 11 25 14 20 27 11-14-14-15 11 10 19-21 20 14 18 21 35-40 12 10 ; HT456
13 24 13 11 16-19 11 12 11 13 11 30 16 9-9 11 11 25 14 20 30 14-16-17-17 10 11 19-21 15 12 18 21 30-34 11 10 10 8 15-15 8 11 10 8 13 10 0 22-24 18 11 12 12 19 7 12 22 18 13 13 12 14 11 11 11 12 33 15 8 15 12 23 26 19 13 12 12 11 12 9 12 11 10 11 12 30 10 13 18 14 11 10 19 16 19 12 23 13 12 16 23 14 21 18 12 13 17 9 12 11 ; HT784
14 25 16 10 11-14 12 12 11 13 11 29 16 9-9 11 11 23 14 20 32 12-12-15-16 12 11 19-23 17 16 17 18 34-36 14 11 ; HT2345

For three of them, we get these results, in single output TXT file:


1. ----------
15 23 15 10 15-15 11 13 11 14 12 32 16 8-10 11 11 25 14 20 27 11-14-14-15 11 10 19-21 20 14 18 21 35-40 12 10 ; HT456
Probability = 100.00% Fitness=46.33 [1.17] I2a2a M223 (for 67+ markers, try level for I-s)
Probability = 0.00% Fitness=21.09 [0.48] I2a1 Isles
Probability = 0.00% Fitness=14.92 [0.37] G1a CTS11562
Probability = 0.00% Fitness=12.10 [0.27] G2a1 Z6552
Probability = 0.00% Fitness=11.89 [0.23] G2a1 Z6552 > L293 >> Z7940
Probability = 0.00% Fitness=11.58 [0.19] E1b1a V38> M329
Probability = 0.00% Fitness=10.94 [0.27] J2a1 PF5087> Z7430
Probability = 0.00% Fitness=10.91 [0.17] I2a1 S21825>> L880 ("Northern France")
Probability = 0.00% Fitness=10.47 [0.28] R2 M479
Probability = 0.00% Fitness=10.41 [0.20] I2c2 Y16419
Probability of unsupported subclade: 54.31%


Level I:
Probability = 19.59% Fitness=46.10 [0.91] I2a2a M223>Z161>L801>CTS6433> L1425
Probability = 10.85% Fitness=45.32 [0.87] I2a2a M223>Z161>L801>CTS6433> SK1258
Probability = 9.01% Fitness=45.08 [0.88] I2a2a M223>Z161>L801>CTS6433> S2364>Y4955
Probability = 3.93% Fitness=43.45 [0.84] I2a2a M223>Z161>L801>CTS6433> S2364>S2361> Z78
Probability = 2.26% Fitness=43.25 [0.85] I2a2a M223>Z161>L801>CTS6433> S2364>ZS20
Probability = 0.05% Fitness=40.86 [0.77] I2a2a M223>Z161>L801>CTS6433 misc
Probability = 0.00% Fitness=38.32 [0.69] I2a2a M223>Z161>L801>CTS6433> S2364>S2361> FGC55856
Probability = 0.00% Fitness=36.01 [0.61] I2a2a M223>Z161>L801> CTS1977>> Y5282> FGC33295
Probability = 0.00% Fitness=35.51 [0.63] I2a2a M223>Z161>L801> CTS1977> BY13707
Probability = 0.00% Fitness=35.38 [0.57] I2a2a M223>Z161>L801> CTS1977>> Y8935



2. ----------
13 24 13 11 16-19 11 12 11 13 11 30 16 9-9 11 11 25 14 20 30 14-16-17-17 10 11 19-21 15 12 18 21 30-34 11 10 10 8 15-15 8 11 10 8 13 10 0 22-24 18 11 12 12 19 7 12 22 18 13 13 12 14 11 11 11 12 33 15 8 15 12 23 26 19 13 12 12 11 12 9 12 11 10 11 12 30 10 13 18 14 11 10 19 16 19 12 23 13 12 16 23 14 21 18 12 13 17 9 12 11 ; HT784
Probability = 100.00% Fitness=58.00 [1.08] E1b1b > V13
Probability = 0.00% Fitness=34.06 [0.75] E1b1b V22
Probability = 0.00% Fitness=31.16 [0.71] E1b1b > V12
Probability = 0.00% Fitness=12.69 [0.20] D1a1a2 F1070
Probability = 0.00% Fitness=19.38 [0.44] E1b1b V1515
Probability = 0.00% Fitness=19.23 [0.41] E1b1b M123>M34> Z841
Probability = 0.00% Fitness=17.66 [0.29] E1b1b L67
Probability = 0.00% Fitness=17.16 [0.36] E1b1b M123>M34> M84
Probability = 0.00% Fitness=14.73 [0.23] E1b1b V68> SK863
Probability = 0.00% Fitness=13.28 [0.27] E1b1a V38>> L485

Level E:
Probability = 100.00% Fitness=81.50 [1.39] E1b1b V13>>Z5017>> Z16988
Probability = 0.00% Fitness=54.19 [0.84] E1b1b V13>>z5017>> BY4526
Probability = 0.00% Fitness=47.15 [0.76] E1b1b V13>>Z5017>> Z17264
Probability = 0.00% Fitness=45.53 [0.79] E1b1b V13>>Z5018> Y145455
Probability = 0.00% Fitness=45.11 [0.77] E1b1b V13 >> Y19509
Probability = 0.00% Fitness=45.19 [0.72] E1b1b V13>>z5017>> S19928
Probability = 0.00% Fitness=44.01 [0.76] E1b1b V13>>Z5018> S2979> Z16659>Y3183
Probability = 0.00% Fitness=43.73 [0.78] E1b1b V13>>Z5017
Probability = 0.00% Fitness=44.01 [0.68] E1b1b V13>>Z5017>> Z17107
Probability = 0.00% Fitness=41.75 [0.75] E1b1b V13>>S7461



3. ----------
14 25 16 10 11-14 12 12 11 13 11 29 16 9-9 11 11 23 14 20 32 12-12-15-16 12 11 19-23 17 16 17 18 34-36 14 11 ; HT2345
Probability = 100.00% Fitness=50.22 [1.19] R1a (for 67+ markers, try level for R1a-s, 70+ subclades)
Probability = 0.00% Fitness=17.35 [0.46] R2 M479
Probability = 0.00% Fitness=14.31 [0.43] Q M346>> M3> M902
Probability = 0.00% Fitness=14.03 [0.40] O2a2 F525
Probability = 0.00% Fitness=13.42 [0.30] R1b (for 67+ markers, try level for R1b-s, 300+ subclades)
Probability = 0.00% Fitness=11.29 [0.31] Q M346>> Z780
Probability = 0.00% Fitness=11.13 [0.20] R1b PH155
Probability = 0.00% Fitness=10.75 [0.24] O2a1 F51
Probability = 0.00% Fitness=10.68 [0.20] E1b1b M123* (xM34)
Probability = 0.00% Fitness=9.94 [0.22] J2a1 Z6063
Probability of unsupported subclade: 23.26%


Level R1a:
Probability = 28.55% Fitness=58.62 [1.03] R1a Z282>M458>> L1029>YP416
Probability = 23.97% Fitness=58.07 [1.13] R1a Z282>M458>> L1029>YP417
Probability = 10.57% Fitness=56.96 [1.08] R1a Z282>M458>> L1029
Probability = 7.04% Fitness=56.49 [0.97] R1a Z282>M458>> L1029> BY30715
Probability = 2.72% Fitness=54.90 [1.02] R1a Z282>M458>> L1029> YP619
Probability = 2.04% Fitness=54.65 [0.93] R1a Z282>M458>> L1029>YP4647
Probability = 1.06% Fitness=53.36 [1.04] R1a Z282>M458>> YP515
Probability = 0.75% Fitness=53.20 [1.01] R1a Z282>M458>> L1029>YP593
Probability = 0.02% Fitness=48.54 [0.94] R1a Z282>M458>> L1029>YP263
Probability = 0.00% Fitness=46.77 [0.82] R1a Z282>M458>> L1029>YP1703


We also offer option to send to our customers generated images with visual statistics of fitting of the haplotype into predicted haplogroup (they are explained throughly in separate article). This means one image for every prediction, which will in our sample mean 6 images, two for every haplotype (one for prediction in General Level, and one for haplogroup-specific Level). Images are in PNG format, about 15-16 kilobytes long in average. We also have the option to send to our customers generated images with visual statistics of 10 most probable subclades for every prediction, for up to 500 haplotypes.

Haplotypes sent to us must be in the same format and in the same file, so that we can manually convert data to appropriate format for all haplotypes together (not for every haplotype individually). We do not accept data which does not meet this requirement. The best would be data to be in Excel file (.XLS), but we can also support TXT or HTML files. Would be good that every haplotype have some kind of ID (although it is not required), which will be used in final output report with prediction, and can be used for naming images, because it will make analysis of results easier to our customers.

We do not use or publish data sent to us for prediction, and we erase data 14 days after job is done (we keep them for 14 days just in case they would be needed to be recalculated or resent again). So, our customers can be sure we keep privacy of their data sent to us.

Beware that our haplogroup predictions are provided 'as-are', without any express or implied warranty. In no event will the authors of NevGen Predictor be held liable for any damages arising from the use of them.

We accept payments through PayPal. Price of our services can be found here:
nevgen.org/NevGen - price of batch prediction.pdf

Our customers can contact us on:

Please, just indicate do you want images to be generated with prediction.



We would also appreciate any donation to NevGen Predictor from other Predictor users, in order to help pay costs of our server and many days of work spent into making it better.
Our PayPal account is the same previous address:


Also, we want to use possible donations and/or batch prediction income to pay Y700 testing in Family Tree DNA for interesting samples from Balkans related to authors of this site.
We are collecting donations for two of them now:
I1-P109 sample from Zvornik, eastern Bosnia.
I2a1a > PH908 sample from Komani tribe, Montenegro.

Here are images generated in our sample:

prediction 1.1

prediction 1.2

prediction 2.1

prediction 2.2

prediction 3.1

prediction 3.2