Change search
ReferencesLink to record
Permanent link

Direct link
Amount of information needed for model choice in Approximate Bayesian Computation
Uppsala University, Disciplinary Domain of Science and Technology, Biology, Department of Ecology and Genetics, Plant Ecology and Evolution.ORCID iD: 0000-0003-3344-6036
Institut National de la Recherche Agronomique (INRA), UMR Agroecologie, Dijon.
Uppsala University, Disciplinary Domain of Science and Technology, Biology, Department of Ecology and Genetics, Plant Ecology and Evolution.
Institut National de la Recherche Agronomique (INRA), UMR Interactions Arbres-Microorganismes, Nancy.
2014 (English)In: PLoS ONE, ISSN 1932-6203, Vol. 9, no 6, e99581- p.Article in journal (Other academic) Published
Abstract [en]

Approximate Bayesian Computation (ABC) has become a popular technique in evolutionary genetics for elucidating population structure and history due to its flexibility. The statistical inference framework has benefited from significant progress in recent years. In population genetics, however, its outcome depends heavily on the amount of information in the dataset, whether that be the level of genetic variation or the number of samples and loci. Here we look at the power to reject a simple constant population size coalescent model in favor of a bottleneck model in datasets of varying quality. Not only is this power dependent on the number of samples and loci, but it also depends strongly on the level of nucleotide diversity in the observed dataset. Whilst overall model choice in an ABC setting is fairly powerful and quite conservative with regard to false positives, detecting weaker bottlenecks is problematic in smaller or less genetically diverse datasets and limits the inferences possible in non-model organism where the amount of information regarding the two models is often limited. Our results show it is important to consider these limitations when performing an ABC analysis and that studies should perform simulations based on the size and nature of the dataset in order to fully assess the power of the study.

Place, publisher, year, edition, pages
2014. Vol. 9, no 6, e99581- p.
Keyword [en]
approximate bayesian computation, population genetics, bottlenecks, summary statistics
National Category
Evolutionary Biology Probability Theory and Statistics
URN: urn:nbn:se:uu:diva-207706DOI: 10.1371/journal.pone.0099581ISI: 000338633900009PubMedID: 24959900OAI: diva2:649279
Available from: 2013-09-18 Created: 2013-09-18 Last updated: 2014-08-16Bibliographically approved
In thesis
1. Neutral and Adaptive Processes Shaping Genetic Variation in Spruce Species
Open this publication in new window or tab >>Neutral and Adaptive Processes Shaping Genetic Variation in Spruce Species
2013 (English)Doctoral thesis, comprehensive summary (Other academic)
Abstract [en]

Population genetic analyses can provide information about both neutral and selective evolutionary processes shaping genetic variation. In this thesis, extensive population genetic methods were used to make inferences about genetic drift and selection in spruce species. In paper I we studied four species from the Qinghai-Tibetan Plateau (QTP): Picea likiangensis, P. purpurea, P. wilsonii and P. schrenkiana. Big differences in estimates of genetic diversity and Ne were observed in the more restricted species, P. schrenkiana, and the other more widely distributed species. Furthermore, P. purpurea appears to be a hybrid between P. likiangensis and P. wilsonii. In paper II we used Approximate Bayesian Computation (ABC) to find that the data support a drastic reduction of Ne in Taiwan spruce around 300-500 kya, in line with evidence from the pollen records. The split from P. wilsonii was dated to between 4-8 mya, around the time that Taiwan was formed. These analyses relied on a small sample size, and so in Paper III we investigated the impact of small datasets on the power to distinguish between models in ABC. We found that when genetic diversity is low there is little power to distinguish between simple coalescent models and this can determine the number of samples and loci required.

In paper IV we studied the relative importance of genetic drift and selection in four spruce species with differing Ne: P. abies, P. glauca, P. jezoensis and P. breweriana. P. breweriana, which has a low Ne, exhibits a low fraction of adaptive substitutions, while P. abies has high Ne and a high fraction of adaptive substitutions. The other two spruce, however, do not support this suggesting other factors a more important. In paper V we find that several SNPs correlate with both a key adaptive trait (budset) and latitude. The expression of one in particular (PoFTL2) correlates with budset and was previously indentified in P. abies. These studies have helped characterise the importance of different population genetic processes in shaping genetic variation in spruce species and has laid some solid groundwork for future studies of spruce.

Place, publisher, year, edition, pages
Uppsala: Acta Universitatis Upsaliensis, 2013. 30 p.
Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology, ISSN 1651-6214 ; 1078
spruce, population genetics, adaptation, evolution, picea, approximate bayesian computation, cline
National Category
Evolutionary Biology
urn:nbn:se:uu:diva-207714 (URN)978-91-554-8760-7 (ISBN)
Public defence
2013-10-31, Lindahlsalen, Evolutionary Biology Centre, Uppsala, 09:30 (English)
Available from: 2013-10-10 Created: 2013-09-18 Last updated: 2014-01-23

Open Access in DiVA

fulltext(907 kB)72 downloads
File information
File name FULLTEXT01.pdfFile size 907 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Other links

Publisher's full textPubMed

Search in DiVA

By author/editor
Stocks, MichaelLascoux, Martin
By organisation
Plant Ecology and Evolution
In the same journal
Evolutionary BiologyProbability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar
Total: 72 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Altmetric score

Total: 260 hits
ReferencesLink to record
Permanent link

Direct link