Summer School on Bayesian Statistics in Freiburg

As in the last years, we organize a course in Bayesian statistics this summer. See invitation below, feel free to distribute widely.

3^rd International Summer School

An Introduction to Bayesian Modeling for Ecologists

in Freiburg (Germany) 28^th July to 2^nd August 2013

 

Motivation

Bayesian inference is an increasingly used statistical framework in ecology, but it is still rarely taught within the standard curriculum. This course provides a practical introduction to Bayesian inference for students, graduate students, and early-stage researchers. Our goal in teaching this course is that you will be able to apply Bayesian methods to a wide range of typical statistical problems in ecological research.

 

Lecturers

Florian Hartig, University of Freiburg

Jörn Pagel, University of Potsdam / Université Montpellier II

Joseph Chipperfield, University of Trier

 

Content

The content of the course will include:

• Introduction to concepts of Bayesian statistics (Priors, Likelihoods, etc.)
• Sampling methods (e.g. Markov Chain Monte Carlo)
• Bayesian modeling and hierarchical Bayesian models
• The Bugs model language and its implementations such as JAGS and OpenBugs

 

The course will consist of lectures, practical exercises (with R and JAGS) and talks on advanced topics in Bayesian statistics. Course language will be English. If necessary, we can provide certificates with ECTS credits from the University of Freiburg. This is a introductory course, but a basic knowledge of conventional statistics as well competency in R or another programming language will be highly beneficial to profit from this course and will also be considered for selecting the participants.

 

Venue and course fees

The course will take place at the “Unihaus”, a small university guest house in the beautiful mountainous surroundings of Freiburg in Southern Germany. There will be a course fee of 280 Euros for PhD students and higher, and a reduced rate of 230 Euros for students, which covers a basic accommodation in dormitories and all meals, except for beverages which have to be purchased on site. Please plan to arrive in Freiburg on the afternoon of Sunday, 28^th July. The course will finish on Friday, 2^nd August, at lunchtime.

 

How to apply

To apply for the course, please send an email with your contact details, a brief summary of your research interests and motivation for attending this course (max 150 words), and an indication of your experience with R/programming/statistics by 20^th May to joern.pagel@uni-potsdam.de. After this date, we will make the initial selection and subsequent applications will be admitted onto the course on a first-come first-served basis until all places are filled.

Reproducibility at Nature

In an editorial in tomorrow’s issue, Nature announces an initiative to increase reproducibility of research results. They state that “From next month, Nature and the Nature research journals will introduce editorial measures to address the problem by improving the consistency and quality of reporting in life-sciences articles”. The announced measures include

To ease the interpretation and improve the reliability of published results we will more systematically ensure that key methodological details are reported, and we will give more space to methods sections. We will examine statistics more closely and encourage authors to be transparent, for example by including their raw data.

[...]

To further increase transparency, we will encourage authors to provide tables of the data behind graphs and figures. This builds on our established data-deposition policy for specific experiments and large data sets. The source data will be made available directly from the figure legend, for easy access.

Commenters under the article point out that this doesn’t solve all problems, but to me these definitely sounds like many steps in the right direction. The other side of the coin, of course, is to provide the incentives to actually reproduce published studies. High-profile journals could probably also contribute here by providing appropriate space, an aspect that I missed in the announcement.

The fish that first walked on land …

is probably *TATA* – the lungfish.

OK, I spoiled the surprise, but if you have any interest in evolution, you should still read this excellent paper on the evolution of the first tetrapods in tomorrow’s issue of Nature. It’s just amazing what the molecular biology colleagues are pulling out of full genome sequencing data in the last years. Certainly also a handy paper for arguing with the occasional creationist that claims we don’t understand the pathways to key innovations in evolutionary history.

Balancing Selection Shared Between Humans and Chimpanzees

The new issue of Science will appear tomorrow, but I still wanted to spend some more words on the current issue, which has clearly been a treat for the ecologically minded reader – not only did it solve the mystery of fairy circles, but there are also two papers on pollination (services) and a quite interesting study that demonstrates the importance of mycorrhiza and other fungi for the accumulation of soil-organic matter in boreal forests (see also a blog post here).

And there is the study I want to highlight here, namely “Multiple Instances of Ancient Balancing Selection Shared Between Humans and Chimpanzees” by Ellen M. Leffler and colleagues, which I found highly interesting. They test for instances of balancing selection, i.e.

adaptation that leads to the persistence of variation in a population or species in the face of stochastic loss by genetic drift.

As the authors explain, it is difficult to search for balancing selection in the genome of one species alone, as

patterns of diversity can occur by chance because of the tremendous variance in coalescence times due to genetic drift alone (14). As an illustration,
under a simple demographic model with no selection, the probability that two human lineages do not coalesce before the split with chimpanzee is on the order of 10−4 (15, 16). Although this probability is small, the human genome is large, and so many such regions could occur by chance.

To overcome this problem, the study used

complete genome sequences from 59 humans from sub-Saharan Africa (Yoruba) (20) and 10 Western chimpanzees (Pan troglodytes verus) (21) in order to identify shared polymorphisms— namely, high-quality orthologous SNPs with identical alleles in the two species

The methods for identification of shared polymorphisms and also the interpretation of the identified genes in the conclusions seemed a bit voodoo to me, but given that I know very little on this field I was happy to give the authors the benefit of the doubt and simply enjoy this very nicely written and clearly motivated paper.

References

Leffler, Ellen M.; Gao, Ziyue; Pfeifer, Susanne; Ségurel, Laure; Auton, Adam; Venn, Oliver; Bowden, Rory; Bontrop, Ronald; Wall, Jeffrey D.; Sella, Guy; Donnelly, Peter; McVean, Gilean & Przeworski, Molly (2013) Multiple Instances of Ancient Balancing Selection Shared Between Humans and Chimpanzees. Science, 339, 1578-1582.

Fairy circles – a mystery solved?

Fairy circles are circular vegetation-free patches with diameters of a few meters that occur in arid regions of Africa, particularly in Namibia. There has been a great deal of discussion about their origin, with hypothesis ranging from termites over modification of the environment by certain plants to purely geological / physical mechanism such as wind turbulences or geological activity. A new paper by Norbert Juergens in Science suggests that the sand termite, Psammotermes allocerus, generates the pattern

through removal of short-lived vegetation that appears after rain, leaving circular barren patches. Because of rapid percolation and lack of evapotranspiration, water is retained within the circles. This process results in the formation of rings of perennial vegetation that facilitate termite survival and locally increase biodiversity. This termite-generated ecosystem persists through prolonged droughts lasting many decades.

Is that the last word spoken on the origin of fairy circles? Looking at the history of this problem, I somehow suspect not, but if so, yet another great conference dinner discussion topic is gone – will we end up having to discuss about the latitudinal gradient all the time?

The future of academic publishing – a new page*

Tomorrow’s issue of Nature has a special section devoted to the future of academic publishing. The topics covered are not particularly surprising (see also my older posts on this subject), but I found the articles really informative and enjoyable to read, in particular “Open access: The true cost of science publishing”, a News Feature by Richard Van Noorden (see picture below), or the comment on OA licensing by John Wilbanks.

What strikes me is that the discussion about “the future of publishing” currently still centers around legal and economic problems rather than on a media transition. Don’t get me wrong, OA is a good thing and should be a top priority for decision makers in research, but a far more interesting question seems to me how communication of research results will be done in 50 years, how and which content will be produced, and how it will be accessed and used. An article by Jason Priem looks at this, but I find there are still a lot of question marks on this particular page.

* Actually, a lot more pages than one, but I just went with the original Nature title

Figure from Van Noorden, Nature, 2013.

Back from Bayes

I just returned from teaching a workshop on Bayesian statistics together with Joe Chipperfield in Göttingen. The workshop was organized by the Biodiversity, Macroecology and Conservation Biogeography group in the aftermath of the Macroecology meeting, so quite a few people stayed over the weekend after the meeting, and it generally turned out to be a really nice and productive course (at least that was my impression).

In case you would have been interested in the course: as in the last years, we’ll organize a summer school on Bayesian statistics this year, the invitation is not out yet but it is practically certain that it will be in Freiburg, Germany, last week of July, so save the date! (the invitation will be posted here)