Ebola in West Africa — a look at recent data

Confirmed cases of Ebola in West Africa in July and August 2014. Red: confirmed accumulated infections, blue: confirmed accumulated deaths.
Confirmed cases of Ebola in West Africa in July and August 2014. Red: confirmed accumulated infections, blue: confirmed accumulated deaths. The vertical count axis is scaled logarithmically. Figure by hoffmann based on data published by the World Health Organization, Creative Commons License.

The World Health Organisation releases so-called Disease Outbreak News (DONs) for a number of infections, including Ebola Virus Disease. The Ebola DONs referring to the current outbreak in West Africa (Guinea, Liberia, Nigeria, and Sierra Leone) show a sustained, approximately exponential increase in the number of confirmed Ebola infections and deaths in (see Figure) during July and August 2014. There are no clear signs of a subsiding of the outbreak. In the past two months, the number of confirmed cases has roughly doubled each month. The death rate is stable at about 50%. Bad.

(See also update of post.)

What is significant? The Global Mean Rank Test may bring the answer

Comparison of the Global Mean Rank test with the established tests SAM and LIMMA. The Global Mean Rank test has consistently a low False Discovery Rate (FDR) and a high True Positive Rate (TPR). The performance of the Mean Rank Test is high and relatively robust against different types of data processing.

In the public perception, statistics is at best boring if not dubious. However, many modern scientific methods generate huge amounts of data. To find important information in these data, new statistical methods are needed. For instance, such methods can tell us which genes or proteins are important for a certain disease amongst the many thousands that form just background noise? The Global Mean Rank Test (“MeanRank” for short) is a statistical test that has been developed for such purposes. It is based on a very simple idea: If an element (e.g. a protein, a gene, etc.) turns up very prominently in several comparable experiments which study a certain phenomenon (e.g. a disease), it is probably “significant” for this phenomenon. The authors have implemented this idea in an efficient way, and they could show that MeanRank performs well with real world problems.

PLOS ONE: Identification of Significant Features by the Global Mean Rank Test.

Surprising features of the Sonic Hedgehog protein

Hedgehog proteins are important “morphogens” that steer embryonic development in concentration-dependent ways. Despite many years of intensive research, the mechanism of morphogen action is still under debate. We have studied properties of ShhN, the actual signaling part of Sonic Hedgehog, by a comprehensive set of computational methods and based on experimentally determined molecular structures. Our work suggests surprising new features of ShhN, namely that ShhN is an enzyme that acts as a cannibalistic peptidase, and that this enzymatic function is switched off by the binding of calcium ions to ShhN. Our computational approach reveals the details of this novel switching mechanism. To test the predicted autodegradation of ShhN, we study in vitro the stability of specific mutants, and we find that these experiments are in agreement with predictions.

The Figure shows the putative catalytic center of the ShhN peptidase in several conformational states (A: comparison of X-ray structures, B: schematic of calcium triggered switch, C: computational simulation shows switching behaviour consistent with X-ray structures).

 

PLOS Computational Biology: Signaling Domain of Sonic Hedgehog as Cannibalistic Calcium-Regulated Zinc-Peptidase.

2.1 billion years old fossils from Gabon — early multicellular organisms?

El Albani et al. describe structures that could be fossils of early multicellular organisms or assemblages: “The Francevillian Formation contains centimeter-sized structures interpreted as organized and spatially discrete populations of colonial organisms living in an oxygenated marine ecosystem.”

via PLOS ONE: The 2.1 Ga Old Francevillian Biota: Biogenicity, Taphonomy and Biodiversity.

Statistical association of gender and career status in life science academia

2014-08-17-notes-bayes-1
Figure: Posterior probability densities of fraction \theta of female scientists (black curve) and professors (red curve). 95% highest density intervals are marked by dashed lines.

In Germany, the number of female professors is much smaller than expected from the number of female students and scientists, and the fraction of women in “grade A” academic positions is considerably lower than in many other European countries (see e.g. the European She Figures ). When I analysed the recently published gender report of my faculty, I saw this for the first time in numbers for my faculty. In fact, there is a strong association between gender and career status. Specfically, the Figure here shows that the fraction of female professors (red line in Figure around \theta=0.25 or 25%) in 2012 is clearly different from the fraction of female scientists (black line distributed around \theta = 0.6 or 60%) at the level of PhD students and postdocs. If you prefer a frequentist test over the Bayesian analysis shown in the Figure: when applying a Fisher’s exact test to a contingency table of  gender (female vs. male) and career status (scientists vs. professors) we obtain a p-value of 0.006 and an odds ratio of 4.48, supporting a significant association of gender and career status.

To study this association, its possible causes and effects, and ways to overcome this association, my faculty is organizing a public panel discussion (in German) on October 10, 2014 in the great lecture hall at the Essen Campus. We have invited several competent scientists from industry and academia for the panel (more about this in a later post).