See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/41381511

Origin of the Jews and the Arabs: Date of their Most Recent Common Ancestor is
Written in their Y-Chromosomes - However, There Were Two of Them

Article  in  Nature Precedings · February 2010

CITATIONS READS

2 8,678

1 author:

Anatole Klyosov
Harvard Medical School
191 PUBLICATIONS   3,124 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

DNA genealogy View project

Biomedicine View project

All content following this page was uploaded by Anatole Klyosov on 28 May 2014.

The user has requested enhancement of the downloaded file.

Origin of the Jews and the Arabs: Date of their Most

Anatole A. Klyosov1

A pattern of Y-chromosomal mutations in 37 and 67 marker haplotypes of the Jews

and the Arabs indicates that their most recent common ancestor in haplogroup J1

(subclade J1e*) and that (a different one) in haplogroup J2 (subclade J2a*) lived

4300±500 years before present (ybp) and 4175±510 ybp, respectively, that is

practically at the same time. Then a split between the Jewish and the Arabic

lineages in both J1 and J2 haplogroups occurred, which is clearly visible on the

respective haplotype trees. The data show that a common ancestor of Cohanim

(Jewish High Priests) of haplogroup J1 lived 1070±170 ybp, while a common

ancestor of Cohanim in haplogroup J2 lived 3300±400 ybp.

______________________________________________________________
1
MIR International, Inc., 36 Walsh Road, Newton, Massachusetts 02459, USA

[email protected]

1
DNA genealogy, the essential goal of which is to describe and to date the origins of

human populations in terms of time spans to their most recent common ancestors, has

evolved with an astonishing speed during the last decade. An application of principles of

chemical kinetics to mutations accumulated in certain markers (loci) of Y-chromosome (a

chosen combination of markers in Y-chromosome is called a haplotype) lays a well

justified ground for the respective chronological calculations. Those, in turn, serve as a

measuring stick with respect to the history of humanity.

Essentially, at the core of the methodology, is a translation of the average number of

mutations per marker in Y-chromosomes in the chosen population into a respective

timespan, that is a coalescence time focusing at the common ancestors’ lifetime. To do

this correctly, the right approach should take into account a contribution of reverse

mutations as well as a degree of symmetry or asymmetry of mutations, which increase or

decrease a number of tandem repeats in the haplotypes loci. Besides, a right choice

should be made for mutation rate constants for the given (or chosen) haplotypes, which

are different for different combinations of markers in haplotypes. It should also

accomplish a separation of lineages descended from different common ancestors in the

chosen population (in a case of mixed populations, which commonly happens in real

situations). A rather comprehensive description of the approach resolving these

complications, and including criteria for composing and analyzing haplotype trees

(examples are shown below) is given in1.

2
 It is well known that many Jews and Arabs belong to haplogroups J1 and J2, as well as

many other haplogroups2-8. For some, apparently, quite arbitrary reasons, haplogroup J1

has been considered as the most likely one to host the principal ancestor of Cohanim, or

Cohens, the Jewish High Priests. Granted, the first studies of the origin of the Cohens did

not resolve haplogroups J into J1 and J29,10, however, the subsequent papers on the

subject have shifted that origin to haplogroup J1, specifically, subclade J1e8. It was

repeatedly claimed that a certain “signature” of alleles in certain markers (DYS 19, 388,

390, 391, 392, and 393), namely 14-16-23-10-11-12, identified the so-called “Cohen

Modal Haplotype” (CMH) in haplogroup J16,8.

Haplogroup J1

In this study we have collected 94 of 37 marker haplotypes of haplogroup J1, all having

the CMH “signature” as above, Jewish and non-Jewish, including several Arabic

haplotypes (ironically, they all had the “Cohen Modal Haplotype”), and composed the

respective haplotype tree, shown in Fig. 1.

Each bar in Fig. 1 represents a haplotype1. The tree shows two rather distinct branches:

one, on the left, which is compact, and contains Jewish haplotypes only and another, on

the right, which is “fluffy”, therefore indicating a significantly “older age” in terms of a

timespan to a common ancestor of the branch. It contains mostly non-Jewish haplotypes,

including all three Arabic haplotypes. All 44 haplotypes of the “older” branch contain

291 mutations in the first 25 markers (0.265 mutations per marker on average), and 543

3
mutations in the “faster” 37 markers from the base (deduced ancestral) haplotype as

follows:

12 23 14 10 13 17 11 16 11 13 11 30 -- 17 8 9 11 11 25 14 20 26 12 14 16 17 -- 11 10 22

22 15 14 18 18 32 35 12 10

This translates into 4100±460 years to a common ancestor of the “older” Arabic-Jewish

branch, as explained in the METHODS section at the end of this article (all other data on

mutations and their calculations are given in that section).

The “younger”, Jewish only, branch has a common ancestor 1125±160 years before

present (ybp) who had the following haplotype

12 23 14 10 13 15 11 16 12 13 11 30 -- 17 8 9 11 11 26 14 21 27 12 14 16 17 -- 11 10 22

22 15 14 20 18 31 35 13 10

These two base haplotypes differ by 5.5 and 10.5 mutations in their 25 and 37 marker

format, respectively, which places their common ancestor to 4300±500 ybp, and makes

the common ancestor of the “older” branch the most recent common ancestor of both the

Jews and the Arabs in both the mixed “older” and the Jewish “younger” branch.

4
 In other words, we see the two lineages derived from a common ancestor who lived

around 4300 ybp, one encompasses mainly Jews, and another – mainly non-Jewish

people, among them the Arabs.

To verify this tentative conclusion and examine the time estimate for the split in finer

detail, we analyzed 67 marker haplotypes, available from 34 individuals within the

preceding population (see Fig. 1). The respective haplotype tree is shown in Fig. 2.

The 67 marker haplotype tree has provided even better separation of the “younger” (on

the right) and the “older” (on the left) branches (Fig. 2). All 17 haplotypes of the

“younger” branch belong to Jews, and 11 out of 17 haplotypes of the “older” branch were

non-Jewish, including the only one Arabic haplotype reported in the 67 marker format.

The base haplotypes in the 67 marker format were as follows:

“Older branch”:

12 23 14 10 13 17 11 16 11 13 11 30 -- 17 8 9 11 11 25 14 20 26 12 14 16 17 -- 11 10 22

22 15 14 18 18 32 35 12 10 -- 11 8 15 16 8 11 10 8 11 9 12 21 22 18 10 12 12 15 8 12 25

21 13 12 11 14 12 12 12 11

and “Younger” branch:

5
12 23 14 10 13 15 11 16 12 13 11 30 17 -- 8 9 11 11 26 14 21 27 12 14 16 17 -- 11 10 22

22 15 14 20 18 31 35 13 10 -- 11 8 15 16 8 11 10 8 11 9 12 21 22 17 10 12 12 15 8 12 24

21 13 12 12 14 12 12 12 11

These haplotypes are identical with the above (for the respective branches) at the first

37 markers, and provide an extension with 30 more markers. A common ancestor for the

“older”, mixed branch lived 4225±545 years before present, which is within the margin

of error with that for the 37 marker dataset (4100±460 ybp).

The “younger”, Jewish only branch on the right-hand side in Fig. 2 had a common

ancestor who lived 1020±170 ybp. Again, this is within the margin of error with

1125±160 ybp obtained for the 37 marker dataset. The data is quite reliable, and

calculations employing the 25, 37, and all 67 marker haplotypes in Fig. 2 gave 900±195,

1150±180, and 1000±140 ybp, respectively, to the lifetime of a common ancestor of the

Jewish population (the “younger” branch on the right in Fig. 2), all of them having the

“Cohen Modal Haplotype”.

As in the preceding (37 marker) example, we see the two lineages derived from a

common ancestor who lived around 4300 ybp, and he was the most recent common

ancestor of both the Jews and the Arabs. He lived – taking into account both the 37 and

67 marker datasets 4300±500 years before present.

6
 His Arabic descendants continued his direct lineage, while his Jewish descendants went

through a population bottleneck at 1070±170 years ago (this is also called a genetic drift)

and expanded, beginning at about the 10th century AD into what is now called the “Cohen

Modal Haplotype” lineage.

In fact this should be called the “Abraham Modal Haplotype” when it is exhibited in

Jews and Arabs.

Data, obtained by us in11, provides additional support to this suggestion. In that study,

haplotypes of actual Cohens (Cohanim) of haplogroup J1e*-P58*, collected by Hammer,

Behar, Skorecki et al8 were considered, and it was found that their common ancestor

lived 1075±130 years before present11. Three ancestral haplotypes of J1e* Cohanim,

identified in the study11 were – in the 22 marker format available (the last shown allele is

number 37th in the above extended haplotypes):

12 23 14 10 13 15 11 16 12 13 11 30 – 17 8 9 11 11 26 14 21 26 -- 10

12 23 14 10 13 15 11 16 12 13 11 30 – 17 8 9 11 11 26 14 21 27 -- 10

12 23 14 10 13 15 11 16 11 13 11 30 – 17 8 9 11 11 26 14 21 26 -- 10

One can see that the middle one exactly matches the first 21 markers, along with the

37th one, of the “younger” branch in Figs. 1 and 2. Two other base (ancestral) haplotypes

differ from it by only one and two mutations (marked in bold), and a common ancestor of

all the three base haplotypes lived in the range of 625-875 ybp11.

7
 To sum it up, the most recent common ancestor of Jews and Arabs of haplogroup J1

(subclade J1e) lived 4300±500 years ago, and he had the “J1 Abraham Modal

Haplotype”, former “Cohen Modal Haplotype” signature. From him a split occurred

between the Jewish and the Arabic lineages in haplogroup J1 (J1e*). The split is clearly

visible on the haplotype tree in the 37 and the 67 marker format (Figs. 1 and 2).

1070±170 years before present, in the 10th century (plus-minus a century), a line of

Cohanim, haplogroup J1e*, split from that ancestral lineage.

The most recent common ancestor of the Cohanim J1 (J1e*) lineage had a haplotype

which still contained the “Abraham” six-marker signature 14-16-23-10-10-12 (see

above), however, his haplotype already contained six mutations from the initial haplotype

in that lineage.

Those six mutations were all inherited by Cohanim in haplogroup J1, and many of them

continue to carry those mutations to the present time.

Haplogroup J2

We have collected 181 and 131 of 37 and 67 marker haplotypes, respectively, of

haplogroup J2. All haplotypes were obtained from Jewish and Arabic Projects, such as

Arabic World DNA, J2-Arab Project, Sharifs DNA Project, Jewish Heritage Project,

8
 Cohen Project, Sephardic Heritage project, and from YSearch database. The respective

haplotype trees are shown in Figs. 3 and 4.

J2a4b1
550 ybp
J2a4b

675 ybp
450 ybp

Arabs,
J2a
Arabs
and
Turks
The main branch,
J2a
Jewish
branch,
J2a

Most of the branches in Fig. 3 contain either only Jewish haplotypes, or only Arabic

ones, and cannot be employed for the main purpose of this work. Besides, some branches

were not only mono-ethnic, but were too recent in terms of timespans to their common

9
ancestors (see Fig. 3). Clearly, those branches represent rather recent lineages, and are

not related to the genealogical split between the Arabs and the Jews. There is only one

branch on the tree which appears appropriate for the analysis of the split. This is a double

branch in the lower right-hand side of the tree, one half of which is represented with

mainly Jewish haplotypes, and another with mainly Arabic ones. It is marked as “The

main branch” on the tree. All haplotypes on the branch belong to J2a haplogroup. The

branch has 39 haplotypes, however, eight of them form a sub-branch which is only 450

years “old” and its haplotypes noticeably differ from those of the rest of the branch.

Clearly, it is a recent formation, and contains only Arabic haplotypes.

The rest of the branch containing 31 haplotypes (11 were Arabic and 20 were Jewish

haplotypes) has a common – Jewish and Arabic – base (ancestral) haplotype of

12 23 15 10 14 17 11 15 12 13 11 29 -- 15 8 9 11 11 24 15 21 31 12 13 16 17 -- 10 10 19

23 16 14 18 18 36 37 12 9 – 11 7 14 15 8 11 10 8 11 9 12 17 17 14 10 12 12 15 9 12 22

21 14 12 11 14 11 12 12 12

A common ancestor of the joint Jewish-Arabic branch lived 4375±530 ybp, after which

the joint population split.

The 37 marker haplotype tree (Fig. 4) contains more haplotypes compared to the 67

haplotype tree. The tree is essentially the same, with just a few branches switched their

positions.

10
 J2a4b

J2a4b1

550 ybp

The main branch

The base 37 marker haplotype of the main branch (containing 37 haplotypes) is

identical to that in the first 37 markers of the base 67 marker haplotype (see above). A

common ancestor of the joint Jewish-Arabic branch, according to the 37 marker

haplotypes, lived 3975±480 ybp, after which the population split.

Hence, a common ancestor of both Jewish and Arabic lineages in haplogroup J2a-

M410* lived – as an average of the timespans estimated from the 67 and 37 marker base

haplotypes - 4175±510 years before present.

11
 In other words, the time of the split between the Jewish and the Arabic lineages

occurred at the same time in J1 (4300±500 ybp) and J2 haplogroups.

It is remarkable that the base haplotype for descendants of Cohanim of the former

Russian Empire (Russia, Ukraine, Belarus, Lithuania, and Poland) of haplogroup J2a-

M410*, determined in paper11

12 23 15 10 14 17 11 15 12 13 11 29 -- 15 8 9 11 11 24 15 21 32 -- 9

(the last allele belongs to the 37th marker in the extended haplotypes above) is practically

identical with the respective 22 markets of the above extended haplotypes. The only

mutation (in bold) is not a surprise, since a common ancestor of those Cohanim lived

325±190 years ago, and a common ancestor of the 37 and 67 marker haplotypes above

lived 4175±510 years ago, almost four thousand years earlier. It is remarkable, though,

that the joint Jewish-Arabic common ancestral haplotype has retained its pattern ca. 4,000

years after the lifetime of the founder of both the Jewish and the Arabic original

lineages.

There is yet additional proof of the significance of the above ancestral haplotype in the

J2 Cohanim lineage. This proof was presented in the Cohanim haplotypes themselves.

President of the International Jewish Association of Cohanim, Latin America, Mr.

Mashuah Pereira Cohen, has provided me with a list of twenty one verified traditional

12
Cohanim surnames from both Ashkenazi and Sephardic communities, which are – by

their status – in good terms with the Association, along with their 12-marker haplotypes,

obtained from FTDNA projects. Among them were Cohanim-Ashkenazim from Russia,

Belarus, Poland, Lithuania, Hungary, Germany, Austria, Spain, Portugal, Ireland, as well

as Cohanim-Sephardim from Spain, Portugal, Brazil, Morocco and the Netherlands. Their

last names, according to the Association, include Cohen, Kahan, Mazer, Kaplan, Katz,

Cowan, Coyne, HaKohen, Cohen Pereira, Cohen Rodrigues, Machado (Levy), Shapiro,

Ben Ezra, Levy, Pereira Cunha, Garfinkel, Kagan, Kovacs, Kohn, Kohen, Cunha. Their

haplotype tree is shown in Fig. 5.

Seven haplotypes of the Cohanim-Sephardim had 14 mutations from their base

(ancestral) haplotype

12 23 15 10 13 18 11 15 12 13 11 29

which gives 2500±700 years from their common ancestor. 14 haplotypes of the Cohanim-

Ashkenazim had 18 mutations from their base haplotype

12 23 15 10 14 17 11 15 12 13 11 29

which gives 1575±400 years from their common ancestor. One can see that there are only

two mutations between the base Cohanim Ashkenazim and Sephardim haplotypes (shown

in bold), which translates into 2500 years of mutational difference between their common

13
ancestors, and places their common ancestor to 3300±400 years before present. It fits into

the pattern emerging here, that the 12-marker base J2 Cohanim (Ashkenazim) haplotype

is identical with the first 12 alleles in the haplotype of the common ancestor of the Jews

and the Arabs (shown above in the 67 marker format). .

To sum this section up, the most recent common ancestor of the Jews and the Arabs of

haplogroup J2a-M410* lived 4175±510 years ago, and he had the “J2 Cohanim”

signature in his haplotype. Again, it is rather “J2 Abraham Modal Haplotype”. From

him a split occurred between the Jewish and the Arabic lineages in haplogroup J2. The

split is clearly visible on the haplotype tree in the 37 and the 67 marker format (Figs. 3

and 4). Around 3300 years before present, in the middle of the 2nd millennium BC, a line

of Cohanim, haplogroup J2a*, split from that lineage and continued with the initial

ancestral haplotype. Around 2500 years before present, in times of the Babylonian and

the Assyrian exiles, the J2 Cohanim-Sephardim lineage split, and continued with the two

mutations at DYS385a,b (13-18 compared to the initial 14-17) in the ancestral haplotype.

The rest of J2 Cohanim continued with the same ancestral haplotype, and passed a

population bottleneck (a genetic drift has occurred) in the middle of the 1st millennium

AD.

This concurs with well-known historical data, according to which some 3300 years ago

an institute of Cohanim was established, with its inheritance status. The Sephardim split

from the main body of the Jewish population in the middle of the 1st millennium BC; and

after the destruction of the Temple, in the middle of the 1st millennium AD, a new Jewish

14
community began to form in Europe, and gave rise to Ashkenazim, the “Jews of

Germany”.

Why there were two “Abrahams”, whatever were their real names, one in haplogroup

J1 and another in J2, with very similar, if not identical times of their lives? Furthermore,

how is it that each of the “Abraham” lineages hosts the Cohanim, albeit the Cohanim of

haplogroup J2 have their common ancestor at about 3300 years before present, while a

common ancestor of the Cohanim of haplogroup J1 has appeared only around 1000 years

before present?

A plausible explanation can be offered. Obviously, there were predecessors of the Jews

and the Arabs in both haplogroup J1 and J2. They were practically the same people, and

were called the Bedouins (or we call them Bedouins now). 4200±500 years before

present they split, on the grounds of some apparently very serious reason, which likely

had a religious, that is a cultural and spiritual connotation. The split was, judging from

the sharpness of parting of their DNA genealogical lineages, quite a decisive one.

Naturally, the split occurred not along the haplogroups J1 and J2, but across them. That

was how the Jews and the Arabs had acquired both J1 and J2 haplogroups. The story of

Abraham and his siblings, Ishmael, and Isaac and Jacob, the patriarchs of the Arabs and

the Jews, respectively, was told and re-told by the Arabs and the Jews of all the

haplogroups. Hence, it is reasonable to believe that each haplogroup which was involved

in the separation process, would have had its own “Abraham”, who lived about 4,200

years ago.

15
 (If some haplogroup does not show such a split involving both the Arabs and the Jews

and going back to about 4,200 ybp and earlier, then either there were no Jews and Arabs

with such a haplogroup in those times, or they were not involved in the split).

To verify this hypothesis, I have composed a 25 marker haplotype tree (not shown here)

from 163 haplotypes of haplogroup R1b1b2, which included 108 haplotype of Jews and

155 haplotypes of non-Jews (mostly Europeans, including several Arabs, since only a

very limited number of Arabic R1b1b2 haplotypes are available in databases). The

Jewish haplotypes did not mix with the European haplotypes, and they form a number of

different branches, while all the Arabic haplotypes were located only in the European

branches. Hence, there was no a split of the Jewish and the Arabic haplotypes in

haplogroup R1b1b2, though a common ancestor of the Jews in R1b1b2 lived 5650±710

years before present. However, those proto-Jews of R1b1b2 have not been involved in

the alleged religious/cultural dispute with the emerging Arabs.

METHODS

Analysis of mutations and their rates. Principal methodology was described in1.

Haplotype trees were constructed using PHYLIP, the Phylogeny Inference Package

program12, as was explained in detail in1,13. A “comb” around the wheel, a “trunk”, in

haplotype trees identifies “base” haplotypes, identical to each other and carrying no

mutations compared to their ancestral haplotypes (e.g., in Fig. 5). The farther the

16
haplotypes lay from the wheel, the more mutations they carry compared to the base

haplotype and the older the respective branch.

The “base” haplotypes are the ancestral haplotypes in an ideal case. However, since

those haplotypes often are deduced ones, it would be inaccurate to call them “ancestral”.

Hence, “base” haplotypes.

The timespans to a common ancestor were calculated using both “logarithmic” and

“linear” approaches. The logarithmic method is based on the assumption (which largely,

or practically always holds true) that a transition of the base haplotypes into mutated ones

is described by the first-order kinetics:

B = Aekt,

that is

ln(B/A) = kt

where:

B = a total number of haplotypes in a set,

A = a number of unchanged (identical, not mutated) base haplotypes in the set,

k = an average mutation rate,

t = a number of generations to a common ancestor.

17
 One can see that the logarithmic method does not consider mutations in haplotypes;

only mutated and non-mutated (base) haplotypes are considered.

Mutation-counting methods are all based on accumulation of mutations in haplotypes

over time. They include the “linear”, “quadratic” (ASD) and “permutational” methods1.

In this paper only the “linear” method is employed. In this method a total number of

mutations in a set of haplotypes is counted, an average number of mutations per marker is

calculated, a correction for back mutations is introduced, either numerically1, or using a

handy table1, and a time span to a common ancestor is calculated, either using the Table1

or applying the respective mutation rates. In other words, it is described by the following

equation

n/N/μ = t

where n is a number of mutations in all N haplotypes in the given series of haplotype, μ is

an average mutation rate per haplotype per generation, and t is a number of generations to

a common ancestor, corrected for back mutations.

Average mutation rates employed in this paper, are1:

For 12 marker haplotypes – 0.022 mutations per haplotype, 0.00183 mutations per

marker,

18
for 25 marker haplotypes – 0.046 mutations per haplotype, 0.00183 mutations per

marker,

for 37 marker haplotypes – 0.090 mutations per haplotype, 0.00243 mutations per

marker,

for 67 marker haplotypes – 0.145 mutations per haplotype, 0.00216 mutations per

marker.

These mutation rates were calibrated employing 25 years per generation. Time spans to

a common ancestor were calculated typically employing the first 25 markers in

haplotypes; these markers normally give the most reproducible data. In some cases

calculations have been conducted using also 37 and 67 markers just to illustrate

reproducibility of calculations.

Error margin for time spans to common ancestors are calculated as described in1, and

correspond to the 95% confidence interval.

Details of the calculations.

Haplogroup J1.

1) Fig. 1, 37 marker haplotype tree, the “older” branch of 44 haplotypes, 291 mutations in

the first 25 marker haplotypes give 291/44/25 = 0.265 mutations per marker,

0.265/0.00183 = 145 generations to a common ancestor without correction for back

mutations, or 169 generations with the correction, that is 169x25 = 4225 years to a

common ancestor. 291 mutations determine1 the margin of error equal to 5.86% for the

19
average number of mutations per marker, hence, 0.265±0.016 mutations per marker, and

to 11.59% for the time span to a common ancestor, hence, 4225±490 years to a common

ancestor .

2) Fig. 1, 37 marker haplotype tree, the “older” branch of 44 haplotypes, 543 mutations in

all 37 marker haplotypes give 543/44/37 = 0.334 mutations per marker, 0.334/0.00243 =

137 generations to a common ancestor without correction for back mutations, or 159

generations with the correction, that is 3975 years to a common ancestor. 543 mutations

determine the margin of error equal to 4.29% for the average number of mutations per

marker, hence, 0.334±0.014 mutations per marker, and to 10.88% for the time span to a

common ancestor, hence, 3975±430 years to a common ancestor.

An average of 4225±490 and 3975±430 ybp gives 4100±460 years to a common

ancestor of the “older” branch in the 37 marker haplotype tree of the Jewish and the

Arabic haplotypes.

3) Fig. 1, 37 marker haplotype tree, the “younger” branch of 50 haplotypes, 98 mutations

in the first 25 marker haplotypes give 98/50/25 = 0.078±0.008 mutations per marker,

0.0784/0.00183 = 43 generations to a common ancestor without correction for back

mutations, or 45 generations with the correction, that is 1125±160 years to a common

ancestor.

4) Fig. 2, 67 marker haplotype tree, the “older” branch of 17 haplotypes, 118 mutations in

the first 25 marker haplotypes give 118/17/25 = 0.278±0.026 mutations per marker,

0.278/0.00183 = 152 generations to a common ancestor without correction for back

20
mutations, or 179 generations with the correction, that is 4475±610 years to a common

ancestor.

5) Fig. 2, 67 marker haplotype tree, the “older” branch of 17 haplotypes, 210 mutations in

the first 37 marker haplotypes give 210/17/37 = 0.334±0.023 mutations per marker,

0.334/0.00243 = 137 generations to a common ancestor without correction for back

mutations, or 159 generations with the correction, that is 3975±480 years to a common

ancestor.

An average of 4475±610 and 3975±480 ybp gives 4225±545 years to a common

ancestor of the “younger” branch in the 67 marker haplotype tree of the Jewish and the

Arabic haplotypes.

6) Fig. 2, 67 marker haplotype tree, the “younger” branch of 17 haplotypes, 27 mutations

in the first 25 marker haplotypes give 27/17/25 = 0.0635±0.012 mutations per marker,

0.0635/0.00183 = 35 generations to a common ancestor without correction for back

mutations, or 36 generations with the correction, that is 900±195 years to a common

ancestor.

7) Fig. 2, 67 marker haplotype tree, the “younger” branch of 17 haplotypes, 68 mutations

in the first 37 marker haplotypes give 68/17/37 = 0.108±0.013 mutations per marker,

0.108/0.00243 = 44 generations to a common ancestor without correction for back

mutations, or 46 generations with the correction, that is 1150±180 years to a common

ancestor.

21
8) Fig. 2, 67 marker haplotype tree, the “younger” branch of 17 haplotypes, 93 mutations

in all 67 marker haplotypes give 93/17/67 = 0.0817±0.008 mutations per marker,

0.0817/0.00216 = 38 generations to a common ancestor without correction for back

mutations, or 40 generations with the correction, that is 1000±140 years to a common

ancestor.

An average of 900±195, 1150±180, and 1000±140 ybp gives 1020±170 years to a

common ancestor of the “younger” branch in the 67 marker haplotype tree of the Jewish

and the Arabic haplotypes.

An average of 4100±460 and 4225±545 ybp gives 4160±500 years to a common

ancestor of the “older” branch in the 37 and 67 marker haplotype tree of the Jewish and

the Arabic haplotypes.

An average of 1125±160 and 1020±170 ybp gives 1070±170 years to a common

ancestor of the “younger” branch in the 37 and 67 marker haplotype tree of the Jewish

and the Arabic haplotypes.

Base haplotypes of the “older” and “younger” branches differ by 5.5 and 10.5

mutations on the first 25 and 37 marker haplotypes. This results in a difference between

the two common ancestors of 3400 years, and places their common ancestor to 4300±500

years before present. This is the common ancestor of the “older” branch, within the

margin of error.

22
Haplogroup J2.

1) Fig. 3, 67 marker haplotype tree, the main (subclade J2a) branch of 31 haplotypes, 211

mutations in the first 25 marker haplotypes give 211/31/25 = 0.272±0.019 mutations per

marker, 0.272/0.00183 = 149 generations to a common ancestor without correction for

back mutations, or 175 generations with the correction, that is 4375±530 years to a

common ancestor.

2) Fig. 4, 37 marker haplotype tree, the main (subclade J2a) branch of 37 haplotypes, 231

mutations in the first 25-marker haplotypes give 231/37/25 = 0.250±0.016 mutations per

marker, 0.250/0.00183 = 137 generations to a common ancestor without correction for

back mutations, or 159 generations with the correction, that is 3975±480 years to a

common ancestor.

An average of 4375±530 and 3975±480 give 4175±510 years to a common ancestor of

the “younger” branch in the 67-marker haplotype tree of the Jewish and the Arabic

haplotypes.

1. Klyosov, A.A. DNA Genealogy, mutation rates, and some historical evidence

written in the Y-chromosome: I. Basic principles and the method. J. Genetic

Genealogy, 5, 186-216 (2009).

23
2. Behar, D.M., Thomas, M.G., Skorecki, K., Hammer, M.F., Bulygina, E.,

Rosengarten, D., Jones, A.L., Held, K., Moses, V., Goldstein, D., Bradman, N

and Weale, M.E. Multiple origins of Ashkenazi Levites: Y chromosome evidence

for both Near Eastern and European ancestries. Am. J. Hum. Genet. 73, 768 – 779

(2003).

3. Behar, D.M., Garrigan, D., Kaplan, M.E., Mobasher, Z., Rosengarten, D., Karafet,

T.M., Quintana-Murci, L., Oster, H., Skorecki, K. and Hammer, M.F.

Contrasting patterns of Y chromosome variation in Ashkenazi Jewish and host

non-Jewish European populations. Hum. Genet. 114, 354-365 (2004).

4. Nebel, A., Filon, D., Weiss, D.A., Weale, M., Faerman, M., Oppenheim, A. and

Thomas, M. High-resolution Y chromosome haplotypes of Israeli and Palestinian

Arabs reveal geographic substructure and substantial overlap with haplotypes of

Jews. Hum. Genet. 107, 630-641 (2000).

5. Nebel, A., Filon, D., Brinkmann, B., Majumder, P.P., Faerman, M. and

Oppenheim, A. The Y chromosome pool of Jews as part of the genetic landscape

of the Middle East. Am. J. Hum. Genet. 69, 1095-1112 (2001).

6. Levy-Coffman, E. A mosaic of people: the Jewish story and a reassessment of the

DNA evidence. J. Genetic Genealogy 1, 12-33 (2005)

7. Klyosov, AA. Origin of the Jews via DNA genealogy. Proceedings of the Russian

Academy of DNA Genealogy (ISSN 1942-7484), 1, No. 1, 54 – 232 (2008).

8. Hammer MF, Behar DM, Karafet TM, Mendez FL, Hallmark B, Erez T,

Zhivotovsky LA, Rosset S, Skorecki K. Extended Y chromosome haplotypes

24
 resolve multiple and unique lineages of the Jewish pristhood. Hum. Genet., 126,

707-717 (2009).

9. Hammer, M.F., Skorecki, K., Selig, S., Blazer, S., Rappaport, B., Bradman, R.,

Bradman, N., Warburton, P.J., Ismajlowicz, M. Y chromosomes of Jewish Priests.

Nature 385, 32 (1997).

10. Thomas, M.G., Skorecki, K., Ben-Ami, H., Parfitt, T., Bradman, N. and

Goldstein, D.B. Origins of Old Testament priests. Nature 394, 138-140 (1998).

11. Klyosov, A.A. A comment on the paper: Extended Y chromosome haplotypes

resolve multiple and unique lineages of the Jewish Priesthood. Human Genetics,

126, 719-724 (2009)..

12. Felsenstein, J. Phylip, the Phylogeny Inference Package. PHYLIP, version 3.6.

Department of Genome Sciences, University of Washington, Seattle (2005).

13. Klyosov, A.A. DNA Genealogy, mutation rates, and some historical evidences

written in Y-chromosome. II. Walking the map. J. Genetic Genealogy, 5, 217-256

(2009).

Additional Resources:

Sharifs DNA Project, http://www.familytreedna.com/public/sharifs/default.aspx

The Jewish R1b Project, http://www.familytreedna.com/public/JewishR1b/default.aspx

J2-Arab Project, http://www.familytreedna.com/public/J2-Arab/default.aspx

Cohen DNA Project, http://www.familytreedna.com/public/Cohen/default.aspx

Jewish DNA Project,

25
http://www.familytreedna.com/public/JewishDNAProject/default.aspx

Jewish Ukraine West Project,

http://www.familytreedna.com/public/Jewish_Ukraine_West/default.aspx

JewishGen Scandinavia DNA Project,

http://www.familytreedna.com/public/ScandinaviaSIG/default.aspx

Sephardic Heritage DNA Project,

http://www.familytreedna.com/public/Sephardic_heritage/default.aspx

Ashkenazi-Levite DNA Project, http://www.familytreedna.com/public/Ashkenazi-

Levite/index.aspx?fixed_columns

YSearch, a Free Public Service from Family Tree DNA, http://www.ysearch.org/

Acknowledgements The author thanks Mr. Mashuah Pereira Cohen, President of the

International Jewish Association of Cohanim, Latin America, for providing haplotypes of

Cohanim, and Mr. Andrew McEachern for valuable discussions in preparations of the

manuscript.

Author Information The author declares no competing financial interest.

FIGURE LEGENDS

Figure 1. 37 marker haplotype tree for the “Cohen Modal Haplotype”, haplogroup

J1. This 94 haplotype tree was composed of haplotypes obtained from YSearch database

and private “Cohen Haplotype” projects. The tight, compact, hence “young” branch on

the left encompasses the Jewish haplotypes only. The fluffy” and much more extended,

therefore “old” branch on the right contains haplotypes, two-thirds of which belong to

26
people claiming non-Jewish origin. Haplotypes 200, 201 and 202s belong to the Arabs

and haplotype 202s belongs to Arab-Sayid/Sharif (Sharifs DNA Project).

Figure 2. 67 marker 34 haplotype tree for the “Cohen Modal Haplotype”,

haplogroup J1. For explanations see the legend to Figure 1.

Figure 3. 67 marker 131 haplotype tree composed of the Arabic (haplotype

numbers below 100) and Jewish (haplotype numbers 100 and higher) haplotypes of

J2 haplogroup. All haplotypes were collected from sources indicated in the text. Some

subclades are indicated on the graph, as well as some timespans to common ancestors for

“young” branches. Branches of subclades J2a4b and J2a4b1 contain only Jewish

haplotypes, with the respective times to their common ancestors of 1200±210 and

700±150 years before present (NOTE for the publisher: the figure does not contain

labels; they are added in the text format, see above)

Figure 4. The 37 marker 181 haplotype tree composed of the Arabic (haplotype

numbers below 100) and Jewish (haplotype numbers 100 and higher) haplotypes of

J2 haplogroup. The 450-ybp Arabic branch in Fig. 3 became more numerous 550-ybp

branch in Fig. 4, that is essentially the same within the margin of error. For further

explanations see legend to Figure 3 (NOTE for the publisher: the figure does not contain

labels; they are added in the text format, see above) .

27
Figure 5. The 12-marker 21 haplotype tree composed of the Cohanim J2 haplotypes,

provided by the International Cohanim Association (President, Mr. Mashuah Pereira

Cohen).

28
View publication stats