03/05/2016
1
03/05/2016
03/05/2016
SOURCE ARE
1
Historical data available
Genetic data available
COLONIZED AREA
Population history (& genetic data)
Past evolutionary and demographic processes have left traces in the genetic variation - analyzing them we attempt to reconstruct evolutionary history of populations
Studying population history = modelling
- Selection of the most appropriate model (evolutionary scenario)
- Estimation of parameters (e.g. time of events, number of founders, duration of bottlenecks, population size, mutation rate)
Description of recent invasions (invasion genetics) Description of older history (phylogeography)
3
03/05/2016
Inferring population history - ABC modelling
We have observed data (e.g. microsatellite genotypes) We know genetic variation and structure
We would like to know which demographic processes and how and when have created such an observed data = population evolutionary history
Why is ABC approach useful in modelling population history?
It allows to deal with much more complex models with many parameters and a lot of complex data (many samples, populations, genetic loci, sequences)
and hence models much more realistic
Population; African population Non-African populatk _Nature Reviews | Genet
Approximate Bayesian Computation
■ model choice and parameter estimation
■ exact LIKELIHOOD function is intractable in complex situations and can be bypassed (approximated) by a SIMILARITY MEASURE between many simulated (under various models) and a single real (observed) data
■ data SIMULATION under various models
■ COMPARISON of simulated and observed data - model choice
■ Acoording to the most supported model we can ESTIMATE VALUES of its parameters - parameter estimation
4
Decreasing of dimensionality
SIMULATED
_JI_y
OBSERVED DATASET
DATA^pS
VERSUS
SUMMARY STATISTICS
VERSUS
Comparison of simulated and real dataset to infer probability of various models (evolutionary scenarios of population history)
kolonizace3_PCA_l_2_5700
m
i 1 1 1 i 1 1 1 i 1 1 1 i 1 * 1 i * * 1 i * 1 * i 1 * 1 i * * * i * 1 * i 1 * * i
-6 -5-4-3-2-10 1 2 3 4
P.C.1 ( 60.0%)
03/05/2016
■ ■  L'lHL! I. - ili-<«-LK.-iiL-s ^.i[T:.l.tonn
Approximate Bayesian Computation in Population Genetics Mark A. Beaumont,*"1 Wenyang Zhang+ and David J, Balding1
* School of Animal and Microbial Sciences, The University of Reading, Whiteknighls, Reading RG6 6Aj, United IGngdon 'Institute oj ."VJi'1';,j.',-;;^;.0. i        1";.       \. { n/vpfi/i", "f K,Jr,-i. IJjHtrrf'uri:. for:! i 11 iSl. ■! r:r;-'<\ Kingdom rind ''Ijsfvirtmsnf    i  j..1'. '!..:::':. I <     j;. V ,':,.>■■.■' .'it',
St. Man's Cerrnfm\ Norfolli Plmv. London W2 //¥.. Uuih'd Kingdom Manuscript received March 22, 2002 .Vfvpifl |i>r | ml ihi ii ii in < >t i'ilii i 1. i<"i'2
ABSTRACT
V .   f.r f.n     i m iUH I. t   .pp r    ifr.n.   I' i* r*. in T.Tt,i il ir.ln n .   r r. if.   r-.-r ri - imn-i in
suvtisti<"s' 1 In1 method is suited In OMiiples. pr'ilileins tli.U ,e,"ts<- n jji -| ;ii la i l< -i i lie net lis. v\\< w \ w\-.\ id i sis i k-vl'j|iL-(l ] ii i Ills set 1111-; I r. ■ nil- i . mi 111 m I" 11 -| n-1111 ■. 11 1111 ■ | -i ■ i 11 ■! i h'.i 111 111 n n i i >f a parameter, such n tit m. an ,-r .1- r.-.ir,   nr.     it.* if f.i  iimil,-d ..irh.nr   tplinl LL  l.h.-. .1 ral   iliiirr- Thi* i* t l-.i ■.-■H
I iv lii iiii:; .i 11   .il-ln ir.n r; ^ro'.i' in i>] si.....I.n; ;l | i.ir. n i n -i; i" e.ili u-s < m sum il.il ri I sum man si. it tsi ks. .iihI i lien
-■iili'-lnuliii.^ i Lie i il">:-erveil Miium.irv smilmlis iiilh tin r< ; ;i ■ n ■! i - ■ |' i- ii i ■ 11 1 \ \- 11 k i 11- " I i «>i ill I lies mairv oi Llae achai liases nE LtaeesuLi st.ilistical inEeL vjkl" i.itli 111 ■ - . i .n 11'i n. n 1.1 n. il .1 liciei uv oJ mei 111 id;i h.nrd 'jn uifT.rrur. ' i Hi n *    '  I  . '.Hnr.i',;: -i ih* m. it. .H r rh     fhr ninv.n-   f :t .m *i. f-   .r    mi rrun. .Ih
niir:-i.ih-i |   .in. in llie sui.iliJ.Hii m sup. s<j   hi   11 ■ ■   lar;.;e .....111 >■=■ i   <>l l l i.i i-^.i. 11 < ■ pa i.unel els lli.it arise in
population gene tics problems can be handled without difficult!'. Simulation results indicate computational and si.ui.lii\il i ll'Linno ili.il luiiip.uvs lavnrablv with tli.i.;- <>] alternative methods pivvinLislv proposed in the literature. We aki i o:ni|>are i lie rrkilt'.i el lii lem i i il 111 It: rniee:. i ihl ,iu i eil n sin;.; in;11 Inn Is based <jh siun niari -.i.ui-ties vtaIi lli'ise ubiained du eelh Jrcini tin- il.u.i m-ilii: M< - Ml ■
NEW APPROACH Approximate Bayesian Computation (ABC) Beaumont et al. 2002, Genetics
- estimations of parameters
- useful for model choice among various scenarios applied on the same data
- the likelihood criterion is replaced by a similarity criterion between simulated & observed datasets
- measured by a distance between summary statistics computed on both datasets
'Approximate Bayesian Computation' topic in Web of Science ,Sm
in genetics, evolutionary biology and ecology
50 52
0   0 112^55
OHfNm^Ln^NcooiOHfNfM^
OOOOOOOOOO-i—It—It—It—It—I
ooooooooooooooo rsirsirsirsirsirsirsirsirsirsirsirsirsirsirsi
6
ABC approach used successfully for description of recent invasion scenarios
Estoup &Clegg 2003, Molecular Ecology: Zosterops lateralis, Pacific islands Estoup et al. 2004, Evolution: Bufomarinus, Australia Pascual et al. 2007, Molecular Ecology: Drosophila subobscura, invasion
Figure 1. Worldwide routes of invasion of Harmonia axyridis. Most likely scenario of invasions into eastern North America (EN A), western North America (WNA), South America (5A), Europe (EU) and Africa (AF) by Harmonia axyridis, deduced from analyses based on approximate Bayesian computation. For each outbreak, the arrow indicates the most likely invasion pathway and the associated posterior probability value (P), with 95% confidence intervals in brackets. Years of first observation of invasive populations are indicated. Initially collected from the native area in 1982, the European biocontrol strain (Ebc; blue arrow] was used for biocontrol efforts in Europe and South America. Introductions to North America from the native area (green arrows) may have involved releases for blQccntrol efforts._
■ Lombaert et al. 2010, PLoS ONE: Harmonia axyridis, invasion through the Atlantic and subsequently to the whole world
03/05/2016
Software
Table 3. Software incorporating ABC.
		
Software	Keywords and Features	Reference
OIY-ABC	Software for fit of genetic data to complex situations. Comparison of competing models. Parameter estimation Computation of bias and precision measures for a given model and known parameters values.	153)
ABC R package	Several ABC algorithms for performing parameter estimation and model selection. Nonlinear heteroscedastic regression methods for ABC. Cross-validation tool.	1541
ABC-SysBio	Python package. Parameter inference and model selection for dynamical systems. Combines ABC rejection sampler, ABC SMC for parameter inference, and ABC SMC for model selection. Compatible with models written in Systems Biology Markup Language (SBML). Deterministic and stochastic models.	!SS)
ABC toolbox	Open source programs for various ABC algorithms including rejection sampling, MCMC without likelihood, a particle-based sampler, and ABC-GLM. Compatibility with most simulation and summary statistics computation programs.	156)
msBayes	Open source software package consisting of several C and R programs that are run with a Perl "front-end " Hierarchical coalescent models. Population genetic data from multiple co-distributed species.	[57]
PopABC	Software package for inference of the pattern of demographic divergence. Coalescent simulation. Bayesian model choice.	158)
ONeSAMP	Web-based program to estimate the effective population size from a sample of microsatellite genotypes. Estimates of effective population see, together with 95% credible limits.	[591
ABC4F	Software for estimation of F-statistics for dominant data.	[60)
2BAD	Two-event Bayesian ADmixture. Software allowing up to two independent admixture events with up to three parental populations. Estimation of several parameters (admixture, effective sizes, etc.). Comparison of pairs of admixture models.	[61]
dot 1 a 1371 /journal .pcbL 1002803. t003		
Sunnákeret al. 2013, PLOS Computational Biology
SOFTWARE
Melhmls ill EvologyimtlEmhilim 21)13,4,6K4-6S7 dor: 10.1111/2041-210X. 12050
APPLICATION
EasvABC: performing efficient approximate Bayesian computation sampling schemes using R
Franck Jabot*, Thierry Faure and Nicolas Dumoulin
Irstea, URUSC Laboratoire d'ingenierie dessystemes complexes, 24 avenue des Landais BP 20085, Aubiere F-63172. France
abc-Csilléry et al. 2011, Methods in Ecology and Evolution Bez GUI:
SimCoal - simulator + ABC regression - Anderson et al. 2005 msBayes - simulator + ABC regression - Hickerson et al. 2007
SGUI:
ONeSAMP - ABC rejection - jen jedna Wright-Fisher populace - Tallmon et al. 2004
popABC - ABC rejection - Lopes et al. 2009
8
Cornuet er at. BMC ßioinformoöß 2010, 11:401 http^^^ruW.bjom(}C|c(>ritral.i:om.il47l-2l 05/11/4Q1
RESEARCH ARTICLE
(bmc
Bioinformatics
Open Access
Inference on population history and model checking using DNA sequence and microsatellite data with the software DIYABC (vl.O)
Jean-Mane Cürnuet1, orpine Ravigne", Amaud E
APPLICATIONS NOTE
DIYABC v2.0: a software to make approximate Bayesian computation inferences about population history using single nucleotide polymorphism, DNA sequence and microsatellite data
Jean-Marie Coniuet1. Piene Pudlo1 : \ Julien Veyssier1 * \ Alexandre Delme-Gaicia1 \ Ma-tliien Gautiei1 , Raphael Leblois1 , Jean-Michel Mann" , and Aniaud Estoup1
' Iura. UMRJ063 Cbgp. Montpellier. France.: UniverMte Montpellier 2. l*MR CMRS 5149.13M. Montpellier. France 1 Institut .te Biologie Coinpiuaiioimelle iIBC). 95 nie de la Oalera. 34095 Montpellier. France. * CNRS-UM:. Institut d Biologie Couiputationiielle. LIRMM. Motupelliei. France
no simple software solution => inaccessible to most biologists
BUT NOW    Do It Yourself: I : software
allows to infer populaton history using the ABC approach
(Cornuet et al. 2008, 2010, 2014)
DIYABC
$ DIYABC 2.0.4 (19-03-2014)
File Help
üjNew MSS   djNew SNP    g^Open     bJ Save    (ijjsave alt
IYABC
Version 2 beta
A   computer  software   to maki Inference on population evolutiO' history using genetic data rosatellites, DNA sequences and )  obtained from population
J
http://www.montpellier.inrs.fr/CBGP/dig3bc/
2.0.4
New Microsat/Sequence project
New SNP project
Open project
03/05/2016
Genetic data
Sequences SNPs
Genotypes
J__J		L_ ll.......
jIjLI		A
ABC works in 3 steps
1. SIMULATION STEP: a very large reference table is produced and recorded
prior parameter distributions
scenario
mutation model
summary statistics (e.g. number of alleles, expected heterozygosity.fst)
the most time-consuming step
based on the genealogical tree of sampled genes and coalescent theory
SIMULATION STEP
Diswpafflm*lffvalj«
Fra~i pricr cislribiJ a-is
Sirriu a-e jertfic ca-a acxj'cirg to scenario end mutalMi model	
	rtpMi ß, limes J
Compute sumniwys1(5:islic!;	
Hcco'd xrsTitlt-i cn;l SLnma".fs:j; clic valjesinarEFerencelahe'ile
REJECTION STEP
anc simlated sjrnrnar, slstislics
le'an simLlaled rata ie's dnsesl :n unserved dale
ESTIMATION STEP		
	Esltaal* posterior p* otiaMlty sc«rsri33 3j ogiüi; 'cgrtssisn	
	Esumate posltflot dlstflbitfon or p5rsrrc:cr sy leca 1 ntar rüg'cssioi	
Cornuet et al. 2008, Bioinformatics
10
03/05/2016
11
SOURCE REGION
DIYABC works in 3 steps
1. SIMULATION STEP: a very large reference table is produced and recorded
2. REJECTION STEP: only the simulated data closest to the observed dataset are retained
based on Euclidian distances in
multidimensional space of summary statistics
SIMULATION STEP
Diswpafflm*lffvalj«
Fra~i pricr cislribiJ a-is
Sirriu a-e jertfic ca-a acxj'cirg to scenario end mifalMi model	
	repMl b, limes J
Compute summary stalistks	
Hcco'd xrsTitlt-i cn;l SLnma".fs:j; clic val jes in □ reFerpnee lah e 'lie
REJECTION STEP
anc si-nLlded sjrnrnarj slstislics
dnsesl :o observed dale
ESTIMATION STEP		
	Esltaal* posterior piotaWSl> sc«rsri33 3j ogiili; 'cgrcssisn	
	Esiimgte posltflot dlstflbitfon or p5rsrrc:cr sy leca 1 ntar rtg'cssioi	
13
Comparison of our observed dataset with simulated ones and inferring posterior distributions of scenarios
kolonizace3_PCA_l_2_5700
03/05/2016
DIYABC works in 3 steps
1. SIMULATION STEP: a very large reference table is produced and recorded
2. REJECTION STEP: only the simulated data closest to the observed dataset are retained
3. ESTIMATION STEP: Estimating posterior distributions of
parameters through a local linear
regression procedure
SIMULATION STEP	
Diaivparamftlwvalü« J        rrom prior flsttlbitfons	
	
^ SirriLj a-p jerr/idtoladCCOTiigId SMnarurinrJmi.latiDn mode	
	IFof each swnarb1|
'      Compute summwy statistics	repeat n, imes I
{Ricnd paramttvr and summary sldcüc I
valjesinarEFerencelahe'ile
REJECTION STEP
C:rfiLledi!ili3i:cs sck'cer cbucvcd □nc simlated 5jmmarv sialics
levari sirnUBled tald daws) :o tinse-jec CrTa
ESTIMATION STEP
-i-i-iatE? pawnor pictawiiv
srswmu:    ogi:1i; •cjressun
EstindrpDstericrcist-ibLlan or pErsrrcicr   leca I near rtg'cssion
Posterior distributions of parameters are estimated according to the most supported scenario
kolonizace3_PCA_l_2_5700
-3.5
-5-4-3-2-10 1 2 3 4
P.C.1 ( 60.0%)
15
03/05/2016
hoto: Jaroslav Červený
Black rat {Rattus rattuš) invasion in Senegal
Konečný et al. 2013, Molecular Ecology
Rattus rattus distribution
16
istoric background
03/05/2016
Genetic data
14 microsatellites (9 - 22 alleles, mean: 14.14) mean allelic richness - 3.06 (range 1.87 - 4.71) mean expected heterozygosity - 0.538 (range 0.323 - 0.762)
both allelic richness and heterozygosity
decreased with longitude
J
18
03/05/2016
		
		
vi		
		
Genetic information
^^^^^^^^^^^
>n of ourrnodels a rs = evo^Jtionary £
Historie information
Formulation of ourrnodels and related parameters = evo^Jtionary scenarios
Model choice Parameters estimation
19
03/05/2016
ABC analysis in four steps - four questions
20
03/05/2016
21
03/05/2016
Comparison of our observed dataset with simulated ones and inferring posterior distributions of scenarios
kolonizace3_PCA_l_2_5700
22