Article | Open | Published:

# Inferring genetic origins and phenotypic traits of George Bähr, the architect of the Dresden Frauenkirche

## Abstract

For historic individuals, the outward appearance and other phenotypic characteristics remain often non-resolved. Unfortunately, images or detailed written sources are only scarcely available in many cases. Attempts to study historic individuals with genetic data so far focused on hypervariable regions of mitochondrial DNA and to some extent on complete mitochondrial genomes. To elucidate the potential of in-solution based genome-wide SNP capture methods - as now widely applied in population genetics - we extracted DNA from the 17th century remains of George Bähr, the architect of the Dresdner Frauenkirche. We were able to identify the remains to be of male origin, showing sufficient DNA damage, deriving from a single person and being thus likely authentic. Furthermore, we were able to show that George Bähr had light skin pigmentation and most likely brown eyes. His genomic DNA furthermore points to a Central European origin. We see this analysis as an example to demonstrate the prospects that new in-solution SNP capture methods can provide for historic cases of forensic interest, using methods well established in ancient DNA (aDNA) research and population genetics.

## Introduction

Advances in modern molecular biology methods and the resulting possibility of extracting genetic information even from ancient specimens, has led to various attempts to reconstruct the genetic legacy of historic individuals. One of the first attempts was made in 2007 on Sven Estridsen, the last Danish Viking king1, who died in 1074 AD. Other attempts in reconstructing the genetic legacy of historic individuals include the cases of Francesco Petrarca2, the identification of the family of Tsar Nicholas II of Russia3, the famous astronomer Nicolas Copernicus4, King Richard III of England5, the Dark Countess6, a proposed blood sample from King Louis XVI king of France7 and most recently the Belgian King Albert I8. In all of these cases, (except for King Louis XVI, where an Exome and shallow WGS approach was performed), either partial mitochondrial information, such as the hypervariable sequence HVS-I, HVS-II or D-Loop of the mitochondria, or a full mitochondral genome were sequenced. While this is sufficient for investigating maternal ancestry lines, it provides little resolution on genetic origin. Foremost, when focusing on mitochondrial data only, there is no information on the paternal ancestry obtained. Additionally, the prediction of disease risks or phenotypic traits such as hair and eye color are not possible when only mitochondrial information is available. While the availability of cheaper sequencing methods and efficient mitochondrial capture techniques enabled researchers to move from targeting control regions to whole mitochondria, the reconstruction of a full high coverage human genome from ancient human remains via high throughput sequencing still remains costly9. In population genetics, where large cohorts of individuals are studied, the cost pressure urged researchers to move on to more cost-efficient and large-scale methods. This has led to the development of specialized in-solution capture methods that target a pre-defined set of SNP positions, as introduced by Haak et al.10,11. In population genetics of ancient human individuals, these methods are now widely applied to recover population specific diagnostic markers. While these approaches target up to 3.7 M SNP positions12 aiming at solely retrieving population diagnostic SNPs in a previously unrivaled resolution, the set of targeted SNPs includes information about various other diagnostic markers as well13. This enables a more detailed phenotypic and disease specific analysis of historic individuals on a much broader level than before.

Unlike for population genetics studies, the focus within forensic case studies is shifted to the identification of individuals and prediction of phenotypic traits. In the case of the historical figure focused on in this study, George Bähr, the main goal was to investigate how much information can be retrieved by modern in-solution SNP capture methods for such studies and whether the approach is generally suitable for characterizing historic individuals. George Bähr is most widely known for his work as architect of several churches and in particular the iconic Dresdner Frauenkirche, an important monument in German history due to its destruction in the last few weeks of the Second World War and its recent reconstruction after the German reunification. Born on the $${15}^{th}$$ of March 1666 in the village of Fürstenwalde, south of Dresden, as the son of a weaver14,15, George Bähr moved to Dresden in 1690 and after several years of work as a carpenter, he was appointed Master Carpenter of the city of Dresden in 170516. During his time there, he was responsible for building both general housing and churches, such as the Orphanage Church in Dresden (1710), the Trinity Church in Schmiedeberg (1713–1716) and several other churches in Forchheim, Königstein, Hohnstein and Kesselsdorf14. In 1722, he began work on his most ambitious project, the Dresdner Frauenkirche. In 1730, he was granted the title of Architect for his service to the city of Dresden over the previous decade, including his work on the Frauenkirche14,15. Unfortunately, Bähr was unable to see this most prominent piece of work in its full glory, as he died following a pulmonary edema at the age of 72 in 1738, five years before the church was finished14. His skeletal remains were initially buried in the Johannis cemetery. However, they were ultimately moved to the crypt of the Frauenkirche in 185414,15,16,17, after the cemetery was desecrated and moved to a different location in the city. Unfortunately, there are no written excerpts or paintings that can be used by historians to gain an impression of the physical and personal appearance of George Bähr. Unlike for other famous architects, such as Matthäus Daniel Pöppelmann of the same century18, there is almost no material other than basic family background available for George Bähr. Even the most complete biographical and historical works, such as the ones by Möllering17, Fischer15 and the most recent biography by Gerlach14, including intensive archival research, did not reveal any more detailed information on him. After the reconstruction of the Dresdner Frauenkirche, from 1990 to 2005, parts of his skeletal remains were found. In order to obtain biological information such as physical appearance and potential risk alleles for genetically inherited diseases from this historic person of interest, we were provided by the George Bähr foundation with bone samples from his skeletal remains. Through in-solution capture, we were able to obtain high coverage genome wide data from George Bähr and used that information to reconstruct his genetic ancestry and phenotypic traits such as skin and eye color. In addition, we found about a dozen risk alleles for medical conditions, including some that might have contributed to his death.

## Results

In total, three independent sequencing experiments were conducted: an initial whole genome shallow shotgun sequencing to determine parameters such as endogenous DNA content, a mitochondrial DNA capture to obtain a full mitochondrial genome and a 390 K SNP capture to obtain high density SNP information on George Bähr. The analysis of the first shallow whole genome shotgun sequencing (WGS), showed a total endogenous DNA content of 62.2%. The mitochondrial DNA capture resulted in a 395 X covered mitochondrial genome, accompanied by two high density SNP in solution capture libraries for population and disease specific SNP detection. On the latter, a mean depth of 28.19 X coverage on the target dataset of 390 K SNPs published in Haak et al.10 was achieved, spanning a total of 317,990 SNPs (with ≈80% target efficiency of the capture). The first aim was to authenticate the analyzed DNA to be of historic origin. In order to authenticate the sequenced fragments, the terminal substitution rates were investigated. Typical double stranded aDNA libraries show cytosine to thymine misincorporations at the 5′ end and guanine to adenine misincorporations at the 3′ ends19,20. These characteristic substitutions accumulate over time and are caused by deamination of cytosine causing miscoding lesions21. As can be seen in Fig. 1, which was created on the intial WGS shallow sequencing run data, up to 7% damage on both 3′ and 5′ ends of the reads can be observed, confirming the presence of ancient DNA.The nuclear 390 K capture libraries were treated with UDG, following a protocol by Briggs et al.20 to remove damage patterns for improved analysis. The same analysis of the (non-UDG treated) mitochondrial capture library showed identical damage patterns as the initial whole genome shotgun library, as well as minimal mitochondrial contamination as described below, increasing the confidence that the samples indeed contain authentic ancient DNA.

In order to confirm whether the sampled individual was male, a molecular sex determination analysis was done on the sequencing data of the 390 K capture. The results as shown in Table 1 show, that the individual was indeed male.

To further exclude a potential contamination of the sampled individual with human DNA from other sources, a mitochondrial contamination test was performed. The estimated mitochondrial contamination was reported to be very low with levels between 0–2%. Quality and authenticity are a major concern in the field of ancient DNA. The last decade has seen a large array of methods to estimate DNA contamination23 as well as reliable criteria for authenticity such as DNA damage patterns21,24. We followed those criteria strictly and used standard methods to estimate mitochondrial and nuclear contamination rates based on heterozygocity of the mitochondrial genome as well as the sex chromosomes. We can show that the DNA extracted from the remains of George Bähr come from a single male individual that shows damage patterns indicative of at least 100 year old DNA21. We therefore conclude the authentic ancient origin of the specimens DNA. A total number of 1,163 known SNPs25 on chromosome X covered at least twice were analyzed, resulting in a very low X-chromosomal contamination estimate of 0.003% with an estimated error of $$7.391683{E}^{-18}$$26.

After the initial verification and authentication process, the paternal and maternal origin of George Bähr was determined. For this purpose, a complete 395 X coverage mitochondrial genome of George Bähr was reconstructed and a quality filtered (q > 30) consensus sequence of his genome was created using schmutzi27. His maternal haplogroup was determined to be H35 using Haplogrep 228, which is a common subclade of haplogroup H in Central Europe29. Furthermore, the Y chromosomal haplogroup of George Bähr was determined to be R1b1a2a1a2-P312 (Table 2). The assigned Y-chromosomal haplogroup is the most common Y chromosome clade of paternal lineages across much of Western Europe, showing a frequency peak in the upper Danube basin and Paris area with declining frequency towards Italy, Iberia, Southern France and the British Isles30.

A principal components analysis, conducted on 317,990 SNP positions, revealed that George Bähr’s SNP profile matches with profiles commonly found in modern central European individuals as shown in Fig. 2. To further explore the relatedness of George Bähr to European populations, an outgroup $${f}_{3}$$ analysis was performed, confirming the initial PCA results, as shown in Fig. 3. To further test whether Africans, South Asians, East Asians, Native Americans and Oceanians share more affinity with George Bähr than with present-day Hungarian, Croatian and French populations, an $${f}_{4}$$ analysis was also performed. The statistics as shown in Table 3 imply that there was no extra ancestry from outside Europe in George Bähr. The results from an unsupervised ADMIXTURE33 analysis also showed no external genetic components in the genome of George Bähr (Fig. 4).

Next, phenotypically interesting SNPs that are considered to be affected by selection were investigated. With the information obtained by the 390 K SNP capture experiment, George Bähr most likely had brown eyes and light skin, as shown in Table 4. This resembles modern individuals from the same area of Germany, where such a phenotype is commonly found today34. Furthermore, Bähr was most likely lactose tolerant as he was heterozygous for the RS4988235 mutation on the LCT gene35,36, again a typical phenotype for central Europeans. The 390K SNP capture panel does not include SNPs that can be used to determine hair color.

To further elucidate what high density SNP capture methods can provide on such specimen, an extensive literature survey was performed using SNPedia and the database mining tool Promethease44. The results of this analysis are shown in detail in Table 5. Several potential candidate mutations were found in George Bähr that are commonly found in modern European populations, such as a variant responsible for the ability to taste bitterness45,46. Interestingly, we also found a large number of SNPs associated with modern diseases like Type-2 diabetes, hypertension and coronary artery disease, which could potentially be related47 to his reported cause of death, pulmonary edema14. Furthermore, a rare variant responsible for age related macular degeneration48 was found to be present in George Bähr’s genome.

## Discussion

Investigating historic individuals based on genetic data still remains challenging and can only shed light on certain aspects of an individual, such as eye and hair color and a set of well established disease markers. Previous studies on historic individuals1,2,3,4,5,6,8 solely focused on the control region of the mitochondrial DNA and in some cases on full mitochondrial genomes. Although this enabled the analysis of at least the maternal relatedness of historic individuals, the analysis of Y-chromosomal data accompanied by a set of autosomal genetic markers permits researchers to recreate a more detailed genetic picture of historic individuals than before.

Within the scope of this project, a complete mtDNA sequence from the skeletal remains of George Bähr and additionally a set of 317,990 SNPs from his autosomes were retrieved. Standard examination of characteristic damage patterns on the initial shotgun screening data and the mitochondrial capture data suggest an ancient origin for the investigated remains. Very low contamination estimates on mitochondrial and Y-Chromosomal level also showed that the retrieved DNA was authentic and no modern human contamination was found. George Bähr’s maternal haplogroup was determined to be H35 and the Y haplogroup was determined to be R1b1a2a1a2-P312, both commonly found in Central European modern populations. Based on phenotypic analysis, George Bähr had brown eyes, light skin pigmentation and was able to digest lactose in adulthood. The population genetic analysis of ancestry with both $${f}_{3}$$ and $${f}_{4}$$ statistics as well as an ADMIXTURE analysis on the set of 317,990 SNPs confirmed previous findings on the mitochondrial level: George Bähr was of Central European descent and shared no additional genetic components with populations outside Europe.

Unfortunately, there is not much of a historic record on George Bähr’s private life. Thus any information that can be obtained on a genetic level that elucidates and enlarges information on him could be important, given his contributions to the history of the city of Dresden. Although George Bähr lived a relatively long life given his time period, his cause of death may have been a pulmonary edema as stated by several authors14. His genetic make up might have contributed to his death given the detected number of variants found related to obesity, diabetes, hypertension and coronary artery disease, which are now widely seen as high- risk factors for such a cause of death63. Although this seems promising in terms of genetic evidence, a direct correlation of such risk factors with an actual cause of death still remains difficult. We see our results as an example of how genome wide information can help to reveal more information on historic individuals for whom scarce or incomplete personal details are available. Written evidence describes that George Bähr’s remains were initially buried at the Johannis cemetery of Dresden and later moved to the crypt of the Frauenkirche in 185414,15,16,17. Unfortunately, given the time period of the reburial and the demolished condition of the Frauenkirche after the Second World War, we cannot exclude entirely the possibility of skeletal mixup. However, our reconstructed genetic profile as well as the historical provenience of the human remains suggest that the analyzed specimens indeed belong to George Bähr.

With the rise of cost-efficient in-solution based SNP capture methods, historic samples can now be investigated in a much more detailed way than ever before. In contrast to previous methods that focused on mitochondria or control regions, the additional information obtained using established SNP capture protocols can provide much more information for researchers or historians to investigate more complex forensic, population genetic and medical questions. Although genetic methods with respect to phenotype predictions made some progress in the last few years, one must keep in mind that direct connections between genotype and phenotype are still challenging. Estimating personal characteristics from genetic data, such as the height or appearance of an individual are in their early stages, as shown for example by Mathieson et al.13. For even more detailed predictions, e.g. facial reconstructions these direct relationships between genotype and phenotype still remain unresolved. Furthermore, the quality of historic genome data is usually inferior to modern genome data and typically introduces additional error sources, rendering statistically profound statements in the context of phenotypic analysis even more complicated.

New and updated capture protocols are incorporating more diagnostic positions and thus provide now even more SNPs for downstream medical and population genetics analysis in the future. We therefore believe that the current SNP capture methods are just the beginning for studies of historic individuals. For example, Mathieson et al.13 stated that larger cohort studies, such as the one conducted by Mallick et al.64, could reveal more and more diagnostically relevant SNPs and associations between SNPs that can hopefully help resolve such questions in more detail in future.

## Methods

### Ancient DNA extraction & Initial Screening

Bone samples were taken under standard precaution and clean conditions from the skeletal remains of George Bähr, which had been placed in the crypt of the Dresdner Frauenkirche. We performed DNA extraction and library preparation steps in clean-room facilities. Bone powder was collected using a dental drill and subsequently DNA was extracted using an established protocol65. We produced indexed libraries using 20 μ aliquot of the generated extract, following the protocol of Meyer et al.66. Additionally, the libraries were enriched for human mitochondrial DNA in a bead based capture protocol using long-range PCR products as bait for hybridization as introduced by Maricic et al.67. We included one negative control for every step of DNA extraction and library preparation to ensure consistency of results. DNA sequencing was performed in an initial screening run for the enriched library pools on the Illumina Genome Analyzer IIx platform with $$2\times 76+7$$ cycles, following the instruction manual for multiplex sequencing (FC-104-400x v4 sequencing chemistry and PE-203-4001 v4 cluster generation kit). In contrast to the manual, the raw reads were aligned to the PhiX 174 reference sequence to obtain training data for a modified base calling application called Ibis68. The reads were then filtered according to their individual indices and went into RAW data processing.

### Nuclear 390 K capture

In the clean room facilities of the Institute for Archeological Sciences in Tübingen, Germany, two further libraries from $$\mathrm{20\ }\mu l$$ extract each were produced in a similar fashion to the screening library, but additionally implementing a UDG and endonuclease $$VIII$$ damage repair treatment20, to remove deaminated bases. The libraries were amplified to reach an amount of about $$\mathrm{1,000\ }ng$$ DNA for each which was subsequently used in an in-solution hybridization capture approach11, targeting a set of $$\mathrm{394,577}$$ SNPs10. DNA sequencing was performed on a HiSeq 2500 with $$2\times 101$$ cycles.

### RAW data processing and authentication

General RAW data processing for the initial shallow whole genome sequencing (WGS), mitochondrial capture dataset and the 390 K SNP capture data was done using the EAGER pipeline69. In all cases, sequence adapters were clipped with Clip&Merge with default settings and the paired end reads were merged respectively. For the initial WGS and the 390 K SNP capture data, the read mapping procedure was performed with BWA70 0.7.15 and reads were mapped against the hg19 human reference genome. For the mitochondrial capture data, reads were mapped against the rCRS reference genome. The CircularMapper approach as implemented in EAGER was used with default settings to increase mapping qualities towards the ends of the utilized mitochondrial reference genome. In all three datasets, WGS, 390 K and mitochondrial capture, DNA damage authentication was performed using our in-house tool DamageProfiler to determine whether characteristic misincorporation patterns of aDNA are present in the investigated datasets21. In addition, the mitochondrial data was tested for potential contamination in the EAGER pipeline using schmutzi27. On the 390 K capture data, the “MoM” estimate from “Method 1” as well as the “new_llh” X-chromosomal authentication method in ANGSD26 was used to quantify potential autosomal contamination on the X chromosome. Furthermore, a molecular sex identification of the remains of George Bähr was performed using the method previously described in Fu et al.22. This approach calculates the number of reads mapping against the target SNPs on the Y chromosome and compares this to the total number of reads mapping against the target SNPs on the autosome. An empirical threshold from the literature (see71) was then used to determine whether the investigated individual was male or female.

### Y-chromosomal analysis

The Y chromosomal haplogroup was determined by examining a set of diagnostic positions on chromosome Y using the ISOGG database version 11.228 (August 19, 2016), utilizing all available positions on the 390 K capture dataset. In order to perform this analysis, the analysis was restricted to reads with a mapping quality higher than 30. Further detailed investigations revealed that mutations separating George Bähr from upstream Y haplogroups such as R1b1a2a1a (see Table 2) are present. For potential haplogroups within the clade investigated R1b1a2a1, R1b1a2a and R1b1a2 (see Table 2) characteristic mutations were found, which made the placement of George Bähr in Y haplogroup R1b1a2a1a2-P312 most likely.

### Population Specific analysis

#### Principal components analysis

A principal components analysis using the smartpca method available in EIGENSOFT31,32 was performed using default parameters and the options lsqproject: YES and numoutlieriter: 0. The investigated sample was projected onto the variation of 777 present-day West Eurasians with $$\mathrm{317,990}$$ SNPs10.

An ADMIXTURE33 analysis was performed after pruning the data for linkage disequilibrium in PLINK72 with the parameters–indep-pairwise 200 25 0.4 retaining 181,529 SNPs of the 390 K capture dataset10. ADMIXTURE was executed with default 5-fold cross validation, varying the number of ancestral populations between K = 2 and K = 15 in bootstraps of 100 with different random seeds. Again, 777 modern West Eurasians and individuals from worldwide representative populations such as Mbuti, Yoruba, Han, Papuan, Karitiana, Eskimo, Uzbek, Amim, Selkup and Kalash were used for the analysis. The lowest cross-validation errors were observed with K = 7.

#### Outgroup $${f}_{3}$$/$${f}_{4}$$ statistics

Additionally, $${f}_{3}$$ statistics of the form $${f}_{3}({\rm{Mbuti}};B\ddot{{\rm{a}}}\mathrm{hr},X)$$ were calculated to test which West Eurasian populations share the most genetic drift with George Bähr. This analysis was performed using ADMIXTOOLS73 with the parameter settings inbreed: YES, computing standard errors with a block jackknife. For the computation of $${f}_{4}$$ statistics of the form $${f}_{4}(\mathrm{Worldwidepopulations},\mathrm{Chimp};\mathrm{Europeans},B\ddot{{\rm{a}}}\mathrm{hr})$$ ADMIXTOOLS73 was applied and again standard errors were computed with a block jackknife.

#### Phenotypic analysis

After uploading a VCF file74 to the respective web service Promethease, a more detailed report is created stating potential causes for diseases as well as phenotypic traits. To ensure that found variants are indeed trustworthy, the IGV tool was used to manually confirm the findings of the method before reporting75.

Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

## References

1. 1.

Dissing, J. et al. The last Viking King: A royal maternity case solved by ancient DNA analysis. Forensic Sci. Int. 166, 21–27 (2007).

2. 2.

Pilli, E. et al. Ancient DNA and forensics genetics: The case of Francesco Petrarca. Forensic Science International: Genetics Supplement Series 1, 469–470 (2008).

3. 3.

Rogaev, E. I. et al. Genomic identification in the historical case of the Nicholas II royal family. Proceedings of the National Academy of Sciences 106, 5258–5263 (2009).

4. 4.

Bogdanowicz, W. et al. Genetic identification of putative remains of the famous astronomer Nicolaus Copernicus. Proceedings of the National Academy of Sciences 106, 12279–12282 (2009).

5. 5.

King, T. E. et al. Identification of the remains of King Richard III. Nat. Commun. 5, 5631 (2014).

6. 6.

Parson, W., Berger, C., Sänger, T. & Lutz-Bonengel, S. Molecular genetic analysis on the remains of the Dark Countess: Revisiting the French Royal family. Forensic Sci. Int. Genet 19, 252–254 (2015).

7. 7.

Olalde, I. et al. Genomic analysis of the blood attributed to Louis XVI (1754-1793), king of France. Sci. Rep 4, 4666 (2014).

8. 8.

Larmuseau, M. H. D. et al. Biohistorical materials and contemporary privacy concerns-the forensic case of King Albert I. Forensic Sci. Int. Genet 24, 202–210 (2016).

9. 9.

Lazaridis, I. et al. Ancient human genomes suggest three ancestral populations for present-day Europeans. Nature 513, 409–413 (2014).

10. 10.

Haak, W. et al. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature 522, 207–211 (2015).

11. 11.

Fu, Q. et al. DNA analysis of an early modern human from Tianyuan Cave, China. Proceedings of the National Academy of Sciences 110, 2223–2227 (2013).

12. 12.

Fu, Q. et al. An early modern human from Romania with a recent Neanderthal ancestor. Nature 524, 216–219 (2015).

13. 13.

Mathieson, I. et al. Genome-wide patterns of selection in 230 ancient Eurasians. Nature 528, 499–503 (2015).

14. 14.

Gerlach, S. George Bähr: der Erbauer der Dresdner Frauenkirche: ein Zeitbild (Böhlau Verlag Köln Weimar, 2005).

15. 15.

Fischer, H. Forschungen zu George Bähr und dem sächsischen Barock I. und II. Teil. Ph.D. thesis, Dresden (1967).

16. 16.

Magirius, H. Die Dresdner Frauenkirche von George Bähr: Entstehung und Bedeutung (Deutscher Verlag für Kunstwissenschaft, 2005).

17. 17.

Möllering, W. George Bähr, ein protestantischer Kirchenbaumeister des Barock (Leipzig, 1933).

18. 18.

Gross, R. et al. Matthäus Daniel Pöppelmann, 1662–1736, und die Architektur der Zeit Augusts des Starken (Verlag der Kunst Dresden, Dresden, 1990).

19. 19.

Briggs, A. W. et al. Patterns of damage in genomic DNA sequences from a Neandertal. Proceedings of the National Academy of Sciences 104, 14616–14621 (2007).

20. 20.

Briggs, A. W. et al. Removal of deaminated cytosines and detection of in vivo methylation in ancient DNA. Nucleic Acids Res 38, e87–e87 (2010).

21. 21.

Sawyer, S., Krause, J., Guschanski, K., Savolainen, V. & Pääbo, S. Temporal patterns of nucleotide misincorporations and DNA fragmentation in ancient DNA. PLoS One 7, e34131 (2012).

22. 22.

Fu, Q. et al. The genetic history of Ice Age Europe. Nature 534 (2016).

23. 23.

Key, F. M., Posth, C., Krause, J., Herbig, A. & Bos, K. I. Mining Metagenomic Data Sets for Ancient DNA: Recommended Protocols for Authentication. Trends Genet. 33, 508–520 (2017).

24. 24.

Stoneking, M. & Krause, J. Learning about human population history from ancient and modern genomes. Nat. Rev. Genet. 12, 603–614 (2011).

25. 25.

Rasmussen, M. et al. An Aboriginal Australian Genome Reveals Separate Human Dispersals into Asia. Science 334, 94–98 (2011).

26. 26.

Korneliussen, T. S., Albrechtsen, A. & Nielsen, R. ANGSD: Analysis of Next Generation Sequencing Data. BMC Bioinformatics 15, 356 (2014).

27. 27.

Renaud, G., Slon, V., Duggan, A. T. & Kelso, J. Schmutzi: estimation of contamination and endogenous mitochondrial consensus calling for ancient DNA. Genome Biol. 16, 224 (2015).

28. 28.

Weissensteiner, H. et al. HaploGrep 2: mitochondrial haplogroup classification in the era of high-throughput sequencing. Nucleic Acids Res 44, W58–W63 (2016).

29. 29.

Behar, D. M. et al. A Copernican Reassessment of the Human Mitochondrial DNA Tree from its Root. Am. J. Hum. Genet. 90, 675–684 (2012).

30. 30.

Myres, N. M. et al. A major Y-chromosome haplogroup R1b Holocene era founder effect in Central and Western Europe. Eur. J. Hum. Genet. 19, 95–101 (2011).

31. 31.

Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis. PLoS Genet. 2, e190 (2006).

32. 32.

Price, A. L. et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat. Genet. 38, 904–909 (2006).

33. 33.

Alexander, D. H., Novembre, J. & Lange, K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 19, 1655–1664 (2009).

34. 34.

Sulem, P. et al. Genetic determinants of hair, eye and skin pigmentation in Europeans. Nat. Genet. 39, 1443–1452 (2007).

35. 35.

Troelsen, J. T. Adult-type hypolactasia and regulation of lactase expression. Biochimica et Biophysica Acta (BBA) - General Subjects 1723, 19–32 (2005).

36. 36.

Wang, Y. et al. The genetically programmed down-regulation of lactase in children. Gastroenterology 114, 1230–1236 (1998).

37. 37.

Bersaglieri, T. et al. Genetic Signatures of Strong Recent Positive Selection at the Lactase Gene. Am. J. Hum. Genet. 74, 1111–1120 (2004).

38. 38.

Enattah, N. S. et al. Identification of a variant associated with adult-type hypolactasia. Nat. Genet. 30, 233–237 (2002).

39. 39.

Soejima, M. & Koda, Y. Population differences of two coding SNPs in pigmentation-related genes SLC24A5 and SLC45A2. Int. J. Legal Med. 121, 36–39 (2006).

40. 40.

Eiberg, H. et al. Blue eye color in humans may be caused by a perfectly associated founder mutation in a regulatory element located within the HERC2 gene inhibiting OCA2 expression. Hum. Genet. 123, 177–187 (2008).

41. 41.

Sturm, R. A. et al. A Single SNP in an Evolutionary Conserved Region within Intron 86 of the HERC2 Gene Determines Human Blue-Brown Eye Color. Am. J. Hum. Genet. 82, 424–431 (2008).

42. 42.

Kimura, R. et al. A Common Variation in EDAR Is a Genetic Determinant of Shovel-Shaped Incisors. Am. J. Hum. Genet. 85, 528–535 (2009).

43. 43.

Fujimoto, A. et al. A scan for genetic determinants of human hair morphology: EDAR is associated with Asian hair thickness. Hum. Mol. Genet. 17, 835–843 (2007).

44. 44.

Cariaso, M. & Lennon, G. SNPedia: a wiki supporting personal genome annotation, interpretation and analysis. Nucleic Acids Res 40, D1308–D1312 (2012).

45. 45.

Kim, U.-K. Positional Cloning of the Human Quantitative Trait Locus Underlying Taste Sensitivity to Phenylthiocarbamide. Science 299, 1221–1225 (2003).

46. 46.

Reed, D. R. et al. The perception of quinine taste intensity is associated with common genetic variants in a bitter receptor cluster on chromosome 12. Hum. Mol. Genet. 19, 4278–4285 (2010).

47. 47.

Ware, L. B. & Matthay, M. A. Acute Pulmonary Edema. N. Engl. J. Med. 353, 2788–2796 (2005).

48. 48.

Despriet, D. D. G. et al. Complement Factor H Polymorphism, Complement Activators, and Risk of Age-Related Macular Degeneration. JAMA 296, 301 (2006).

49. 49.

Ye, S., Willeit, J., Kronenberg, F., Xu, Q. & Kiechl, S. Association of Genetic Variation on Chromosome 9p21 With Susceptibility and Progression of Atherosclerosis. J. Am. Coll. Cardiol. 52, 378–384 (2008).

50. 50.

Cluett, C. et al. The 9p21 Myocardial Infarction Risk Allele Increases Risk of Peripheral Artery Disease in Older People. Circ. Cardiovasc. Genet 2, 347–353 (2009).

51. 51.

Farzaneh-Far, R., Na, B., Schiller, N. B. & Whooley, M. A. Lack of association of chromosome 9p21.3 genotype with cardiovascular structure and function in persons with stable coronary artery disease: The Heart and Soul Study. Atherosclerosis 205, 492–496 (2009).

52. 52.

McPherson, R. et al. A Common Allele on Chromosome 9 Associated with Coronary Heart Disease. Science 316, 1488–1491 (2007).

53. 53.

Wang, W. Y., Zee, R. Y. & Morris, B. J. Association of angiotensin II type 1 receptor gene polymorphism with essential hypertension. Clin. Genet 51, 31–34 (1997).

54. 54.

Bonnardeaux, A. et al. Angiotensin II type 1 receptor gene polymorphisms in human essential hypertension. Hypertension 24, 63–69 (1994).

55. 55.

Hinney, A. et al. Genome Wide Association (GWA) Study for Early Onset Extreme Obesity Supports the Role of Fat Mass and Obesity Associated Gene (FTO) Variants. PLoS One 2, e1361 (2007).

56. 56.

Dina, C. et al. Variation in FTO contributes to childhood obesity and severe adult obesity. Nat. Genet. 39, 724–726 (2007).

57. 57.

Claussnitzer, M. et al. FTO Obesity Variant Circuitry and Adipocyte Browning in Humans. N. Engl. J. Med. 373, 895–907 (2015).

58. 58.

Frayling, T. M. et al. A Common Variant in the FTO Gene Is Associated with Body Mass Index and Predisposes to Childhood and Adult Obesity. Science 316, 889–894 (2007).

59. 59.

Omori, S. et al. Association of CDKAL1, IGF2BP2, CDKN2A/B, HHEX, SLC30A8, and KCNJ11 With Susceptibility to Type 2 Diabetes in a Japanese Population. Diabetes 57, 791–795 (2008).

60. 60.

Burton, P. R. et al. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447, 661–678 (2007).

61. 61.

Do, R. et al. Genetic Variants of FTO Influence Adiposity, Insulin Sensitivity, Leptin Levels, and Resting Metabolic Rate in the Quebec Family Study. Diabetes 57, 1147–1150 (2008).

62. 62.

Cha, S. W. et al. Replication of Genetic Effects of FTO Polymorphisms on BMI in a Korean Population. Obesity 16, 2187–2189 (2008).

63. 63.

Wong, N. D. Epidemiological studies of CHD and the evolution of preventive cardiology. Nat. Rev. Cardiol 11, 276–289 (2014).

64. 64.

Mallick, S. et al. The Simons Genome Diversity Project: 300 genomes from 142 diverse populations. Nature 538, 201–206 (2016).

65. 65.

Rohland, N. & Hofreiter, M. Ancient DNA extraction from bones and teeth. Nat. Protoc. 2, 1756–1762 (2007).

66. 66.

Meyer, M. & Kircher, M. Illumina Sequencing Library Preparation for Highly Multiplexed Target Capture and Sequencing. Cold Spring Harb. Protoc 2010, db.prot5448–pdb.prot5448 (2010).

67. 67.

Maricic, T., Whitten, M. & Pääbo, S. Multiplexed DNA Sequence Capture of Mitochondrial Genomes Using PCR Products. PLoS One 5, e14004 (2010).

68. 68.

Kircher, M., Stenzel, U. & Kelso, J. Improved base calling for the Illumina Genome Analyzer using machine learning strategies. Genome Biol. 10, R83 (2009).

69. 69.

Peltzer, A. et al. EAGER: efficient ancient genome reconstruction. Genome Biol. 17, 60 (2016).

70. 70.

Li, H. & Durbin, R. Fast and accurate long-read alignment with Burrows-Wheeler transform. Bioinformatics 26, 589–595 (2010).

71. 71.

Mittnik, A., Wang, C.-C., Svoboda, J. & Krause, J. A Molecular Approach to the Sexing of the Triple Burial at the Upper Paleolithic Site of Dolní Vĕstonice. PLoS One 11, e0163019 (2016).

72. 72.

Purcell, S. et al. PLINK: A Tool Set for Whole-Genome Association and Population-Based Linkage Analyses. Am. J. Hum. Genet. 81, 559–575 (2007).

73. 73.

Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012).

74. 74.

Danecek, P. et al. The variant call format and VCFtools. Bioinformatics 27, 2156–2158 (2011).

75. 75.

Robinson, J. T. et al. Integrative genomics viewer. Nat. Biotechnol. 29, 24–26 (2011).

## Acknowledgements

We want to thank Kajo Kusen and the George Bähr Stiftung Dresden for their help in getting access to the skeletal remains of George Bähr. Furthermore, we wanted to express our deepest gratitude to Dr. Siegfried Gerlach for his personal help in collecting literature and written information on George Bähr. We wanted to furthermore thank Judith Neukamm for proof reading. C.C.W was supported by the Max Planck Society and Nanqiang Outstanding Young Talents Program of Xiamen University.

## Author information

### Affiliations

1. #### Integrative Transcriptomics, Center for Bioinformatics, University of Tübingen, Tübingen, 72076, Germany

• Alexander Peltzer
•  & Kay Nieselt
2. #### Department of Archaeogenetics, Max Planck Institute for the Science of Human History, Jena, 07745, Germany

• Alexander Peltzer
• , Alissa Mittnik
• , Chuan-Chao Wang
• , Tristan Begg
• , Cosimo Posth
•  & Johannes Krause
3. #### Institute for Archaeological Sciences, University of Tübingen, Tübingen, 72070, Germany

• Alissa Mittnik
• , Cosimo Posth
•  & Johannes Krause
4. #### Department of Anthropology and Ethnology, Xiamen University, Xiamen, 361005, China

• Chuan-Chao Wang

### Contributions

J.K. designed the experiments. J.K. carried out the skeletal sampling. A.M. and C.P. performed the ancient DNA experiments. A.P. and T.B. researched literature and archival data on George Bähr. A.P., A.M. and C.C.W. analyzed the data. A.P. wrote the manuscript with contributions from all co-authors. All authors read and approved the manuscript.

### Competing Interests

The authors declare that they have no competing interests.

### Corresponding author

Correspondence to Johannes Krause.