Studying Modern Human Origins from Neandertal DNA · from Neandertal DNA September 9 2010, Berlin. 2 Studying Modern Human Origins from Neandertal DNA • Closely related species

Max Planck Institute for Evolutionary AnthropologyEvolutionary Genetics

Martin Kircher

Studying Modern Human Origins from Neandertal DNA

September 9 2010, Berlin

2

Studying Modern Human Origins from Neandertal DNA

• Closely related species and our own history

• Ancient DNA

• Importance of how high-throughput sequencing

• Insights from the Neandertal genome

3

Closest relatives

• Tell us:

– How evolution shaped our

• Genome

• Transcriptome

• Proteome

– When things changed

How we adapted to environmental changes

What makes us human!Steiper and Young (2006) Mol. Phyl. & Evol. 41(2):384

Old world monkeys

New world monkeys

~ Lemurs

Apes

4

Closest relatives

• Chimpanzees and Bonobosare our closest living relatives

• ~98.8% genome identical

Chimpanzee Steiper and Young (2006) Mol. Phyl. & Evol. 41(2):384

Old world monkeys

New world monkeys

~ Lemurs

ApesBonobo

5http://www.mnh.si.edu/anthro/humanorigins/index2.htm

Closest relatives

6

• Closest extinct relative

• Lived in Europe and West Asia~400 – 30ka

• Went extinct when modern humans spread in Europe & Asia

• Share ~99.9% of genome

Allow to distinguish recent changes from shared evolutionary history

Neandertal modern human

Neandertals

7

Recent changes ...

appearance of art

world-wide dispersion

agricultural and technological revolution

morphological changes

8

1 cm

Ancient DNA

DNA extract

~ 500 mg

• DNA can be extracted from blood, soft tissues as well as bones, hair and teeth

• Successful for up to 100ka old samples

9

Modern DNA

Ancient DNA

~1μg DNAper gram tissue

~0.0000001-0.001μg DNA per gram tissue

Two types of DNA?

Contamination

10

Minimizing contamination from handling (e.g. Sidron 1253)

Contamination avoidance

EL SIDRÓN (ASTURIAS, SPAIN)

11

Quantifying human contamination

• Neandertal mitochondrial genomes

fall outside of human variation

– 133 fixed differences can be used

as informative sites

Green et al. Cell 2008 / en.wikipedia.org

12

Quantifying human contamination

• Neandertal nuclear genome falls within human genomic variation

• No known fixed differences, other measures possible:

– Triallic sites

– X homozygosity

– Y chromosomal coverage

Green et al. Science 2010

13

Sampling ancient DNA

1.77% 0.86%1.60%

0.27% 0.40% 0.41%

4.17% 4.49% 2.17% 4.10%2.5% 1.7%

14

Neandertal libraries• Screened ~200 DNA extracts from at least 70 fossils from 16 sites

0 20 40 60 80 100 120 140

01

23

4

0 20 40 60 80 100 120 140

01

23

4

% o

f rea

ds

Length

Vi33.16Vi33.25Vi33.26Feld1Mez1Sid1253

Vindija 0.2 – 4.0%El Sidron 0.1 - 0.4%Neander Valley 0.2 - 0.5%Mezmaiskaya 0.8 - 1.5%

15

Feasibility: Neandertal genome

Nov. 2006Aug. 2007May. 2007Aug. 2007Sep. 2007Feb. 2008Jul. 2008

Bone (grams) Sequencing (runs)

20220.20.20.20.2

6.000 (454 GS20)6.000 (454 GS20)4.000 (454 FLX)4.000 (454 FLX)

700 (454 FLX)300 (454 Titanium)20 (Illumina GAII)

(“Proof-of-Principle”) Improved library prep454 FLX upgradeLibrary amplificationLibrary enrichmentTitanium upgradeIllumina/Solexa

1997 Sanger sequencing of hypervariable region of first Neandertal mitochondrial genome

2000 Two additional mitochondrial sequences2005 454 platform becomes available

Neandertal 1x genome project idea is born

16

Neandertal Illumina sequencing

• Improved base caller (Ibis)

• Ancient DNA aware

aligner (ANFO)

• Paired End sequencing:

reconstruction of original molecule

• Deep sequencing: PCR duplicate consensus

17

Improved base calling: Ibis

0 20 40 60 80 100Position in read

0.000.01

0.020.03

Estim

atederrorrate

BustardIbis

GA I (26nt)

GA II (51nt, v1)

GA II (77nt, v2)

GA II (76nt, v3)

GA IIx(101nt, v4)

Perfect 11.3% 39.8% 9.19% 51.52% 62.60%Error 7.1% 2.0% 2.74% 0.89% 0.41%Perfect 23.4% 60.2% 36.58% 58.90% 65.05%Error 5.4% 1.1% 0.73% 0.65% 0.33%

Bustard

Ibis

GA I (26nt)

GA II (51nt, v1)

GA II (77nt, v2)

GA II (76nt, v3)

GA IIx(101nt, v4)

Perfect 11.3% 39.8% 9.19% 51.52% 62.60%Error 7.1% 2.0% 2.74% 0.89% 0.41%Perfect 23.4% 60.2% 36.58% 58.90% 65.05%Error 5.4% 1.1% 0.73% 0.65% 0.33%

Bustard

Ibis

18

Ancient DNA aligner: ANFO

• Short, erroneous and

damaged reads are

difficult to align

• Ancient DNA damage

model, substitutions

and indel aware aligner

• Highr resolution mapping quality: search for second best alignment

Correct alignments: important for downstream analyses

Modified from Briggs et al. NAR 2010

19

Paired End read merging

Position in read0 20 40 60 80 100

Reverse read Forward read

Err

or p

rofil

e

20

Paired End read merging

0 50 100 150

0.0

0.1

0.2

0.3

Insert size

Seq

uenc

ing

erro

r on

read

s [%

]

Average of raw reads (no merging)

Error-informative scores

Position in read0 20 40 60 80 100

Reverse read Forward read

21x error reduction

21

Neandertal Genome ConsortiumMax Planck Institute for Evolutionary Anthropology

* Adrian Briggs * Anne Fischer* Jeffrey Good * Ed Green * Janet Kelso * Johannes Krause* Martin Kircher * Michael Lachmann * Tomislav Maricic* Matthias Meyer * Svante Pääbo * Kay Prüfer * Susan Ptak* Qiaomei Fu * Susanna Rankin * Rigo Schultz * Udo Stenzel * Johann Visagie * Hernan Burbano

Sequencing group at MPI EVA:* Aximu Ayinuer-Petri* Anne Butthof* Barbara Höber * Barbara Höffner * Madlen Siegemund* Antje Weihmann

Museo Nacional de Ciencias Naturales, Madrid* Javier Fortea* Carles LaLueza-Fox* Marco de la Rasilla* Antonio Rosas

Rheinisches Museum/University of Tübingen* Ralf Schmitz

Broad Institute/ MIT * David Reich* Nick Patterson* Chad Nussbaum* Eric Lander

Whitehead Institute* Steve Rozen * Jen Hughes * Helen Skaletsky

LBL* Gavin Crookes

NIH/NHGRI* Jim Mullikin

Uppsala University* Siv Andersson

Oxford University* Daniel Falush

European Bioinformatics Institute (EBI) * Ewan Birney * Paul Flicek * Ben Paten * Michael Hoffmann * Daniel Zerbino

Croatian Academy of Sciences and Arts * Maja Paunovic* Dejana Brajkovic* Jadranka Mauch Lenardic* Zeljko Kucan * Ivan Gusic * Pavao Rudan

Slatkin Lab: UC Berkeley * Hua Chen * Philip Johnson * Anna-Sapfo Malaspinas * Josh Pollack * Montgomery Slatkin * Rasmus Nielsen

U. of Washington, Seattle* Evan Eichler

EMBL, Heidelberg* Peer Bork

CSHL* Greg Hannon * Emily Hodges * Zhenyu Xuan * Michelle Rooks

Cornell University* Andy Clark * Kirk Lohmueller * Carlos Bustamante

454 Life Sciences Inc * Jan Berka * Brian Desany * Lei Du * Michael Egholm * Xavier Gomes * Jerry Irzyk * Clotilde Perbost * Jason Affourtit

22

Science 328, May 2010

2323

Neandertal genome• 454 data from Vindija extracts (206 million reads; 1.4 Gb hominid)

• Illumina data (214 lanes; 2.5 billion raw reads; 4.1Gb hominid)

• ~ 1.5x: coverage for ~63% bases of human genome

24

Browse the genome...

http://neandertal.ensemblgenomes.orghttp://genome.ucsc.edu/Neandertal

25

Time of last common ancestor

x 6.5 ma Neandertal: 12.7% 825,000 yrs

French: 8.0% 520,000 yrsHan: 8.4% 550,000 yrs

Papua: 9.3% 605,000 yrsYoruba: 9.4% 610,000 yrs

San: 10.3% 670,000 yrs

~6.5 Myr

12.7%

5 human HGDP samples,sequenced to ~6-8x (Illumina GAII)

Reference human

26

Catalog of novel featuresfixed in the human genome

• 78 amino acid substitutions

• 45 fixed changes in 5‘ UTRs and223 fixed changes in 3‘ UTRs ofprotein-coding genes

• 1 fixed change in seed region of hsa-mir-1304

....

5 genes with two amino acid changes since Neandertal split:RPTN Epidermal matrix proteinSPG17 Sperm axonemeCAN15 Optic lobe homologTTF1 RNA pol. I termination factorPCD16 Ca-dep. fibroblast adhesion

Burbano et al. Science 2010

27

Adaptive evolution• Adaptive changes spread fast in a population

• Regions will show recent SNPs not known to Neandertal

Haplogroups before selection Regaining diversityAfter selection

28

Positive selection

Top 20 candidate regions:

Type II diabetes

Schizophrenia

Autism

Cleidocranial dysplasia

Down syndrome

Green et al. Science 2010

29

Cleidocranial dysplasia (CCD)• RUNX2 only gene associated with CCD, a skeletal dysplasia

CCD patient Normal rib cage

Ancestral state

30

Neandertals interbreed with modern humans?

• Neandertal mitochondrial DNA outside of known human

variation: No maternal decendents of Neandertals

• Arguments for Neandertal admixture in Europe for

morphological and geographical/temporal reasons: Is

there gene flow detectable in the nuclear genome?

31

Neandertals interbreed with modern humans?

Green et al. Science 2010

13 non-african haplotypes: 10 are shared with Neandertal

32

Neandertals interbreed with modern humans?

1-4% admixture in all (tested) out-of-Africa populations

% Neandertal matching to H2 –% Neandertal matching to H1

-1

0

1

2

3

4

5

6

7

Z-Sc

ore

HGDP01029 (San) HGDP01029 (Yoruba)HGDP01029 (San) HGDP00521 (French)HGDP01029 (San) HGDP00542 (Papuan)HGDP01029 (San) HGDP00778 (Han)HGDP01029 (Yoruba) HGDP00521 (French)HGDP01029 (Yoruba) HGDP00542 (Papuan)HGDP01029 (Yoruba) HGDP00778 (Han)HGDP00521 (French) HGDP00542 (Papuan)HGDP00521 (French) HGDP00778 (Han)HGDP00542 (Papuan) HGDP00778 (Han)

33

What next?

• Higher coverage genome (~20x?)

• Targeted analyses of genomic

regions and candidates

• Functional characterization

of changes!

• Other human forms?

34

Denisovans ...

Contact

PhD or PostDoc at MPI EVA?

– Viola Mittag (Assistant to Svante Pääbo)[email protected]

– Bioinformatics (Janet Kelso)[email protected]

Max Planck Institute for evolutionary AnthropologyEvolutionary GeneticsDeutscher Platz 6D-04103 Leipzig