in ZICHRONNOTE, newsletter of the San Francisco Bay Area Jewish Genealogical Society,
volume 43, number 2 (September 2023)
on pages 16-19
14th century Jews buried in Erfurt, Germany
One of 2022's most momentous genetic studies presented
many types of data from 33 Jews who were buried in
the city of Erfurt in central Germany in the mid-to-late
14th century (Waldman et al., 2022). Some of these
Jews (called "Erfurt-ME") descended from the Jews of
the Rhineland further to the west while others ("Erfurt-EU")
were Knaanic Jews who had lived in Czechia
and Poland and had spoken Slavic. Although both
subgroups had Middle Eastern and Southern European
roots in common, the Knaanic Jews had heavily
intermarried with West Slavic converts to Judaism around
the start of the 13th century, that is, about 5 generations
before the Erfurt samples. Modern Ashkenazic Jews
from Eastern Europe are descended from both groups
but somewhat more from the Rhineland Jews (roughly
a 60:40 ratio) and only a very small amount of ancestry
was contributed later by Sephardic Jewish migrants and
newer Polish converts.
Janice Sellers is a match
The Erfurt sample I will focus on here, because she
matches ZichronNote's former editor and many other
modern people of Ashkenazic descent, has the sample
code I14740 and has been uploaded to GEDmatch as
kit number EJ7716177. This woman died no later than
the year 1398 and her age at death is estimated to have
been between 35 and 50. She had 829,090 SNPs tested,
which provide decent SNP overlaps with modern Ashkenazim.
Different profile than typical for modern Ashkenazi
In GEDmatch's Eurogenes K36 calculator, she scores
a total of 29.14 percent in West Asian elements (adding
up 12.52 percent East Mediterranean, 9.26 percent Near
Eastern, 3.38 percent Arabian, 2.74 percent Armenian,
and 1.24 percent West Caucasian), significantly less
than any modern Ashkenazim. For comparison, as a
full modern Eastern Ashkenazi, my total for those five
elements is 39.83 percent and some Ashkenazim score
several points higher than I do. Her Slavic-associated
components include 4.47 percent Central European,
13.27 percent East Central European (based primarily
on Belarusian, Lithuanian, and Ukrainian reference
samples), and zero percent Eastern European whereas
mine are only 2.75, 5.01, and 0.71 respectively.
Her mtDNA haplogroup, N9a3a1b1, still extant in
today's Eastern Ashkenazi population (GenBank
sample OQ732365 from my data set), is identified in
the study as being "nested within a Central/East Asian
branch" in part because its parent subclade, N9a3a1b, is
found among Turkic-speaking Bashkirs from the Ural
Mountains area (BioProject samples SRS6892124 and
SRS6892238 from the data set of Nagy et al. 2021)
and Mongolic-speaking Daurs from northeastern China
(GenBank sample ON127764 from the data set of Jia
et al. 2022). Simultaneously, she carries higher-than-average
total Khazar-associated elements for a Jew: 2.1
percent East Central Asian (based on Daurs and other
Mongolic and Tungusic peoples) plus 4.04 percent
Volga-Ural (based on Turkic-speaking Chuvash reference
samples) but no amount of Siberian for a total of
6.14 percent. I only score 0.39 percent Volga-Ural and
nothing in the East Central Asian and Siberian elements.
Surviving autosomal DNA Identity by Descent (IBD)
Inside GEDmatch, I found only two valid autosomal
DNA segments of at least 7 centimorgans in length that
are shared between this woman from Erfurt and modern
Ashkenazim: the one I present below and one on chromosome 22
positioned from 34 million to 38 million. In general,
the numbers of valid segments that each medieval Erfurt Jew shares with modern Jews are not large
because only a small minority of their segments have
remained unrecombined in the Jewish population to
the present day. In some cases these segments' survival
can probably be attributed to their location in more
conserved regions of chromosomes or because of later
parental ages in some lineages or because more fathers
than mothers transmitted the segment up to the present
day (less segments inherited from fathers get completely
broken up and recombined by crossovers compared
to those from mothers). Except for a few areas adjacent
to Excess IBD Regions, the outer limit for survival of
intact identical-by-descent segments with lengths of 6
or greater centimorgans is about 20 generations.
I validated these segments by doing triangulation and
parent-child phasing and by ensuring none of them are
located in an Excess IBD Region.
There is no way to determine whether any modern Ashkenazim directly
descend from her or only are related to her through a
common ancestor, such as one of her parents.
The following participants represent only a subset of the total matches I found to this segment:
EJ7716177: 14th-century Erfurt Jewish sample I14740
T788836: Janice M. Sellers = half-Ashkenazi Jewish + half European Christian; tested with Family Tree DNA
T781846: L. Z. G. = Ashkenazi Jewish; tested with Family Tree DNA
A698401: W. Metzger = 3/4th Ashkenazi + 1/4th non-Jewish Italian; tested with AncestryDNA
T577154: David Șloim Marcus = Ashkenazi Jewish; tested with Family Tree DNA
T731722: Shaun Marcus = Ashkenazi Jewish, son of David Șloim Marcus; tested with Family Tree DNA
M411264: Arnold Marcus = Ashkenazi Jewish, son of David Șloim Marcus; tested with 23andMe
M550241: Hillary D. = Ashkenazi Jewish, granddaughter of David Șloim Marcus' brother; tested with 23andMe
DS6892903: modern Polish Jew "Ashkenazi5788" from Iosif Lazaridis' 2014 data set; has relatively low SNP overlaps with the other kits due to chip incompatibilities
Their shared segment on chromosome 11 positioned from 18 million to 21 million:
Chromosome Start Position End Position Centimorgans SNPs
Comparing Kit EJ7716177 (Erfurt I14740) and Kit T788836 (Janice M. Sellers)
11 18,950,092 21,425,793 7.6 540
Comparing Kit EJ7716177 (Erfurt I14740) and Kit T781846 (L. Z. G.)
11 18,787,044 21,425,793 7.8 554
Comparing Kit EJ7716177 (Erfurt I14740) and Kit A698401 (W. Metzger)
11 17,520,339 21,425,793 9.6 630
Comparing Kit EJ7716177 (Erfurt I14740) and Kit T577154 (David Șloim Marcus)
11 18,950,092 21,604,897 7.9 576
Comparing Kit EJ7716177 (Erfurt I14740) and Kit T731722 (Shaun Marcus)
11 18,950,092 21,425,793 7.6 524
Comparing Kit EJ7716177 (Erfurt I14740) and Kit M411264 (Arnold Marcus)
11 18,906,611 21,425,793 7.7 749
Comparing Kit EJ7716177 (Erfurt I14740) and Kit M550241 (Hillary D.)
11 18,973,545 21,425,793 7.6 739
Comparing Kit EJ7716177 (Erfurt I14740) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,278,098 21,183,993 8.1 286
Triangulation
Although triangulation technically refers to a minimum
of 3 matches (the triangle), in my experience at least
4 matches are needed to be sure the matching isn't by
chance. To confirm that a segment is valid, we always
look to find at least four matches to a segment who are not
close relatives and all of whom match each other in every
possible combination with approximately the same start
and end points across the same chromosome. It means
that person A matches B, C, and D on the same segment,
just as person B matches C and D and C matches D. In
addition to that, it is beneficial to find that some close
relatives share this segment. These processes help to
weed out false matches.
Triangulations to establish this cluster of matches is identical-by-descent from a common ancestor:
Chromosome Start Position End Position Centimorgans SNPs
Comparing Kit T788836 (Janice M. Sellers) and Kit T781846 (L. Z. G.)
11 18,796,480 22,803,712 10.3 1,415
Comparing Kit T788836 (Janice M. Sellers) and Kit A698401 (W. Metzger)
11 18,939,666 21,654,066 8 797
Comparing Kit T788836 (Janice M. Sellers) and Kit T577154 (David Șloim Marcus)
11 17,934,383 21,610,969 9.4 1,326
Comparing Kit T788836 (Janice M. Sellers) and Kit T731722 (Shaun Marcus)
11 17,934,383 21,654,066 9.5 1,321
Comparing Kit T788836 (Janice M. Sellers) and Kit M411264 (Arnold Marcus)
11 17,427,363 21,654,066 10.1 1,497
Comparing Kit T788836 (Janice M. Sellers) and Kit M550241 (Hillary D.)
11 16,829,568 23,374,739 13.8 2,002
Comparing Kit T788836 (Janice M. Sellers) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,973,545 22,786,413 10 316
Comparing Kit T781846 (L. Z. G.) and Kit A698401 (W. Metzger)
11 18,562,982 21,587,215 8.5 841
Comparing Kit T781846 (L. Z. G.) and Kit T577154 (David Șloim Marcus)
11 18,787,044 21,574,121 8.1 1,111
Comparing Kit T781846 (L. Z. G.) and Kit T731722 (Shaun Marcus)
11 18,762,983 21,587,215 8.2 1,117
Comparing Kit T781846 (L. Z. G.) and Kit M411264 (Arnold Marcus)
11 18,796,480 21,587,215 8.1 1,064
Comparing Kit T781846 (L. Z. G.) and Kit M550241 (Hillary D.)
11 18,195,348 22,395,192 10.5 1,374
Comparing Kit T781846 (L. Z. G.) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,795,058 21,537,221 8 280
Comparing Kit A698401 (W. Metzger) and Kit T577154 (David Șloim Marcus)
11 18,939,666 21,660,464 8 796
Comparing Kit A698401 (W. Metzger) and Kit T731722 (Shaun Marcus)
11 18,939,666 21,660,464 8 781
Comparing Kit A698401 (W. Metzger) and Kit M411264 (Arnold Marcus)
11 18,882,221 21,660,464 8.1 813
Comparing Kit A698401 (W. Metzger) and Kit M550241 (Hillary D.)
11 18,939,666 21,658,844 8 802
Comparing Kit A698401 (W. Metzger) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,278,098 21,983,447 9.6 304
Comparing Kit T577154 (David Șloim Marcus) and Kit T731722 (Shaun Marcus)
11 198,510 50,668,400 68.9 13,152
Comparing Kit T577154 (David Șloim Marcus) and Kit M411264 (Arnold Marcus)
11 198,510 50,668,400 68.9 12,861
Comparing Kit T577154 (David Șloim Marcus) and Kit M550241 (Hillary D.)
11 12,189,013 21,539,055 15.9 2,705
Comparing Kit T577154 (David Șloim Marcus) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,973,545 21,184,613 7.1 229
Comparing Kit T731722 (Shaun Marcus) and Kit M411264 (Arnold Marcus)
11 198,510 44,890,761 67 11,976
Comparing Kit T731722 (Shaun Marcus) and Kit M550241 (Hillary D.)
11 12,189,013 22,119,239 17 2,810
Comparing Kit T731722 (Shaun Marcus) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,973,545 21,184,613 7.1 222
Comparing Kit M411264 (Arnold Marcus) and Kit M550241 (Hillary D.)
11 12,189,013 22,030,566 16.8 3,693
Comparing Kit M411264 (Arnold Marcus) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,809,542 21,262,591 7.5 325
Comparing Kit M550241 (Hillary D.) and Kit DS6892903 (Polish Jew Ashkenazi5788)
11 18,973,545 21,262,591 7.3 311
Future experimentation with the medieval Erfurt Jewish samples could involve attempting to match them to other medieval European Jews as well as to members of medieval European peoples to whom they were related within genealogically-significant timeframes such as Czechs, Germans, and Khazars. I already was able to find a 7.1 centimorgan segment match with 480 SNPs on chromosome 16 positioned from 18 million to 24 million between sample I14853 (GEDmatch SQ3204446), one of the Erfurt Jews, and sample SB676 (GEDmatch TU3514934), one of the Tzarfati (Northern French) Jews who were killed in Norwich, England in 1190, although this match has not been triangulated or phased. The Jews in Norwich were very genetically and culturally similar to the early and modern Jews from Rhineland Germany.
Postscript
UNESCO names Erfurt's medieval Jewish buildings in Germany as a World Heritage Site.
Erfurt's Jews who were buried in the cemetery and supplied the DNA samples weren't representatives of Erfurt's original Jewish community that had
built the Old Synagogue in the 11th-14th centuries and attended it until the Erfurt massacre and expulsion of 1349, but rather those who moved in
from other towns a few years later.
Acknowledgements
All of the aforementioned living, non-anonymous GEDmatch participants gave me written permission to have their kits represented in this article and they saw drafts of this article before I submitted it.
Thanks to the Erfurt study's co-author Shai Carmi of the Hebrew University of Jerusalem for his feedback on this and similar Erfurt DNA segments.
Thanks also to the Erfurt study's co-author Leo Cooper for feedback on a draft of this article.
Bibliography: |
Kevin Alan Brook is a historian and population geneticist whose most recent book is The Maternal Genetic Lineages of Ashkenazic Jews. |
An article of related interest which links to the article above: "DNA says you’re related to a Viking, a medieval German Jew or a 1700s enslaved African? What a genetic match really means" by Shai Carmi and Harald Ringbauer in The Conversation, March 29, 2024.
An article of related interest which mentions my name in connection to my research above: "From Vikings to Beethoven: what your DNA says about your ancient relatives" by Ewen Callaway in Nature, volume 632, issue 8024 (August 8, 2024): pages 246-249.
Disclosure: Compensated affiliate of genetic testing companies, bookstores, and Nature.com