logotimc
ANR

ANR DATGEN 

Projet ANR-2010-JCJC-1607-01

DATation GENétique à l'ère post-génomique/Genetic dating in the Post-Genomic Era


Objectives

We aim at developing methods of computational statistics and software that characterize 1. the patterns of
genetic variation in populations of living organisms and 2. the evolutionary processes, which generated
the observed patterns.

Characterizing genetic differentiation between populations

Front cover of MBE (March 2013)

Anisotropic patterns
Although the greatest fraction of human genetic variation is explained by variation among individuals, a non-neglected fraction of genetic di fferentiation occurs among populations. A large proportion of this genetic differentiation among populations is explained by geographic distances among populations with more distant populations being more genetically differentiated than nearby populations. We provide statistical techniques to investigate if  the rate at which genetic differentiation increases may be different according to geographic orientations (east-west, north-south, ...).
fig1
Fig1: Main orientations of genetic differentiation for the four continents. Correlation between genetic differentiation and orientational distances computed along the different spatial directions. The major orientations of genetic differentiation are north-south in Europe and Africa,east-west in Asia, but no preferential orientation was found in the Americas. The orientational distances measure the number of kilometers that separated two populations alonga given direction (almost 0 km between Oslo and Florence along the E-W orientation, and almost the crow-fly distance along the N-S direction). Whenthere is one star (*) or more (**, ***), there is significant evidence for an anisotropic pattern.

Software LocalDiff to ascertain non-stationary patterns of genetic differentiation

Non-stationary patterns of isolation by distance arise when genetic differentiation between populations (or between individuals) increases at different rates in different regions of the the species range. Typical patterns include barriers to gene flows, secondary contact zone, corridors for gene flow, or gradients of gene flow across the species range. Using Bayesian kriging, we provide estimates of local genetic differentiation across the species range. Typically measures of differentiation includes Fst and correlation measures.

sweden

Fig 2: Genetic friction for the human population of Sweden. Friction values are larger in the counties with low populationdensities that are located in Northern Sweden. The friction map was computed based on genome-wide SNP data available for hundreds Swedish individuals.

Inferring variation of population sizes based on genetic data

Based on recently developed statistical techniques (Approximate Bayesian Computation), we have shown thatthere was no « speciation bottleneck » at the origin of the modern humans 150,000 years ago. Our analysis refutes a widespread theory inanthropology, which assumes that a demographic bottleneck---concomitant of the penultimate ice age---triggered the speciation of modern humans (see Sjödin et al. 2012).

bottleneck
Fig 3: Schematic overview of the three investigated demographic models. The firstmodel assumes that no bottleneck occurred, whereas the two following models account for a bottleneck during the penultimate ice age. The end of penultimateice age, approximately 130,000 kya, is marked by a dashed line. The map of Africa where potential human refugia are displayed has been excerpted from Lahr and Foley (1998).