Estimating and quantifying uncertainties on level sets using the Vorob'ev expectation and deviation with Gaussian process models

Several methods based on Kriging have recently been proposed for calculating a probability of failure involving costly-to-evaluate functions. A closely related problem is to estimate the set of inputs leading to a response exceeding a given threshold. Now, estimating such a level set—and not solely its volume—and quantifying uncertainties on it are not straightforward. Here we use notions from random set theory to obtain an estimate of the level set, together with a quantification of estimation uncertainty. We give explicit formulae in the Gaussian process set-up and provide a consistency result. We then illustrate how space-filling versus adaptive design strategies may sequentially reduce level set estimation uncertainty.

Corrected kriging update formulae for batch-sequential data assimilation

Recently, a lot of effort has been spent in the efficient computation of kriging predictors when observations are assimilated sequentially. In particular, kriging update formulae enabling significant computational savings were derived. Taking advantage of the previous kriging mean and variance computations helps avoiding a costly matrix inversion when adding one observation to the TeX already available ones. In addition to traditional update formulae taking into account a single new observation, Emery (2009) proposed formulae for the batch-sequential case, i.e. when TeX new observations are simultaneously assimilated. However, the kriging variance and covariance formulae given in Emery (2009) for the batch-sequential case are not correct. In this paper, we fix this issue and establish correct expressions for updated kriging variances and covariances when assimilating observations in parallel. An application in sequential conditional simulation finally shows that coupling update and residual substitution approaches may enable significant speed-ups.

KrigInv: An efficient and user-friendly implementation of batch-sequential inversion strategies based on kriging

Several strategies relying on kriging have recently been proposed for adaptively estimating contour lines and excursion sets of functions under severely limited evaluation budget. The recently released R package KrigInv 3 is presented and offers a sound implementation of various sampling criteria for those kinds of inverse problems. KrigInv is based on the DiceKriging package, and thus benefits from a number of options concerning the underlying kriging models. Six implemented sampling criteria are detailed in a tutorial and illustrated with graphical examples. Different functionalities of KrigInv are gradually explained. Additionally, two recently proposed criteria for batch-sequential inversion are presented, enabling advanced users to distribute function evaluations in parallel on clusters or clouds of machines. Finally, auxiliary problems are discussed. These include the fine tuning of numerical integration and optimization procedures used within the computation and the optimization of the considered criteria.

Comparison of inorganic nitrogen uptake dynamics following snowmelt and at peak biomass in subalpine grasslands

Subalpine grasslands are highly seasonal environments and likely subject to strong variability in nitrogen (N) dynamics. Plants and microbes typically compete for N acquisition during the growing season and particularly at plant peak biomass. During snowmelt, plants could potentially benefit from a decrease in competition by microbes, leading to greater plant N uptake associated with active growth and freeze-thaw cycles restricting microbial growth. In managed subalpine grasslands, we expect these interactions to be influenced by recent changes in agricultural land use, and associated modifications in plant and microbial communities. At several subalpine grasslands in the French Alps, we added pulses of 15N to the soil at the end of snowmelt, allowing us to compare the dynamics of inorganic N uptake in plants and microbes during this period with that previously reported at the peak biomass in July. In all grasslands, while specific shoot N translocation (per g of biomass) of dissolved inorganic nitrogen (DIN) was two to five times greater at snowmelt than at peak biomass, specific microbial DIN uptakes were similar between the two sampling dates. On an area basis, plant communities took more DIN than microbial communities at the end of snowmelt when aboveground plant biomasses were at least two times lower than at peak biomass. Consequently, inorganic N partitioning after snowmelt switches in favor of plant communities, allowing them to support their growing capacities at this period of the year. Seasonal differences in microbial and plant inorganic N-related dynamics were also affected by past (terraced vs. unterraced) rather than current (mown vs. unmown) land use. In terraced grasslands, microbial biomass N remained similar across seasons, whereas in unterraced grasslands, microbial biomass N was higher and microbial C : N lower at the end of snowmelt as compared to peak biomass. Further investigations on microbial community composition and their organic N uptake dynamics are required to better understand the decrease in microbial DIN uptake.

Nuclear markers reveal unexpected genetic variation and a Congolese-Nilotic origin of the Lake Victoria cichlid species flock

Phylogenetic analyses based on mitochondrial (mt) DNA have indicated that the cichlid species flock of the Lake Victoria region is derived from a single ancestral species found in East African rivers, closely related to the ancestor of the Lake Malawi cichlid species flock. The Lake Victoria flock contains ten times less mtDNA variation than the Lake Malawi radiation, consistent with current estimates of the ages of the lakes. We present results of a phylogenetic investigation using nuclear (amplified fragment length polymorphism) markers and a wider coverage of riverine haplochromines. We demonstrate that the Lake Victoria–Edward flock is derived from the morphologically and ecologically diverse cichlid genus Thoracochromis from the Congo and Nile, rather than from the phenotypically conservative East African Astatotilapia. This implies that the ability to express much of the morphological diversity found in the species flock may by far pre–date the origin of the flock. Our data indicate that the nuclear diversity of the Lake Victoria–Edward species flock is similar to that of the Lake Malawi flock, indicating that the genetic diversity is considerably older than the 15 000 years that have passed since the lake began to refill. Most of this variation is manifested in trans–species polymorphisms, indicating very recent cladogenesis from a genetically very diverse founder stock. Our data do not confirm strict monophyly of either of the species flocks, but raise the possibility that these flocks have arisen from hybrid swarms.

Ceramide in Plasma Membrane Repair

The perforation of the plasmalemma by pore-forming toxins causes an influx of Ca2+ and an efflux of cytoplasmic proteins. In order to ensure cellular survival, lesions have to be identified, plugged and removed from the membrane. The Ca2+-driven fusion of lysosomes with the plasma membrane leads to hydrolysis of sphingomyelin by acid sphingomyelinase and a formation of ceramide platforms in the outer leaflet of the lipid bilayer. We propose that the negative curvature, promoted by tighter packing of lipids in the outer layer, leads to an inward vesiculation of the damaged area for its endocytotic uptake and internal degradation. In contrast, the activation of neutral sphingomyelinase triggers the production of ceramide within the inner leaflet of the lipid bilayer, thereby promoting an outward curvature, which enables the cell to shed the membrane-containing toxin pore into the extracellular space. In this process, ceramide is supported by members of the annexin protein family which act as Ca2+ sensors and as membrane fusion agents.

The Nature of Scientific Proof in the Age of Simulations. Is numerical mimicry a third way of establishing truth?

Is numerical mimicry a third way of establishing truth? Kevin Heng received his M.S. and Ph.D. in astrophysics from the Joint Institute for Laboratory Astrophysics (JILA) and the University of Colorado at Boulder. He joined the Institute for Advanced Study in Princeton from 2007 to 2010, first as a Member and later as the Frank & Peggy Taplin Member. From 2010 to 2012 he was a Zwicky Prize Fellow at ETH Z¨urich (the Swiss Federal Institute of Technology). In 2013, he joined the Center for Space and Habitability (CSH) at the University of Bern, Switzerland, as a tenure-track assistant professor, where he leads the Exoplanets and Exoclimes Group. He has worked on, and maintains, a broad range of interests in astrophysics: shocks, extrasolar asteroid belts, planet formation, fluid dynamics, brown dwarfs and exoplanets. He coordinates the Exoclimes Simulation Platform (ESP), an open-source set of theoretical tools designed for studying the basic physics and chemistry of exoplanetary atmospheres and climates (www.exoclime.org). He is involved in the CHEOPS (Characterizing Exoplanet Satellite) space telescope, a mission approved by the European Space Agency (ESA) and led by Switzerland. He spends a fair amount of time humbly learning the lessons gleaned from studying the Earth and Solar System planets, as related to him by atmospheric, climate and planetary scientists. He received a Sigma Xi Grant-in-Aid of Research in 2006

Fast parallel kriging-based stepwise uncertainty reduction with application to the identification of an excursion set

Stepwise uncertainty reduction (SUR) strategies aim at constructing a sequence of points for evaluating a function  f in such a way that the residual uncertainty about a quantity of interest progressively decreases to zero. Using such strategies in the framework of Gaussian process modeling has been shown to be efficient for estimating the volume of excursion of f above a fixed threshold. However, SUR strategies remain cumbersome to use in practice because of their high computational complexity, and the fact that they deliver a single point at each iteration. In this article we introduce several multipoint sampling criteria, allowing the selection of batches of points at which f can be evaluated in parallel. Such criteria are of particular interest when f is costly to evaluate and several CPUs are simultaneously available. We also manage to drastically reduce the computational cost of these strategies through the use of closed form formulas. We illustrate their performances in various numerical experiments, including a nuclear safety test case. Basic notions about kriging, auxiliary problems, complexity calculations, R code, and data are available online as supplementary materials.

Cilengitide treatment of newly diagnosed glioblastoma patients does not alter patterns of progression

The integrin antagonist cilengitide has been explored as an adjunct with anti-angiogenic properties to standard of care temozolomide chemoradiotherapy (TMZ/RT → TMZ) in newly diagnosed glioblastoma. Preclinical data as well as anecdotal clinical observations indicate that anti-angiogenic treatment may result in altered patterns of tumor progression. Using a standardized approach, we analyzed patterns of progression on MRI in 21 patients enrolled onto a phase 2 trial of cilengitide added to TMZ/RT → TMZ in newly diagnosed glioblastoma. Thirty patients from the experimental treatment arm of the EORTC/NCIC pivotal TMZ trial served as a reference. MRIcro software was used to map location and extent of initial preoperative and recurrent tumors on MRI of both groups into the same stereotaxic space which were then analyzed using an automated tool of image analysis. Clinical and outcome data of the cilengitide-treated patients were similar to those of the EORTC/NCIC trial except for a higher proportion of patients with a methylated O(6)-methylguanyl-DNA-methyltransferase gene promoter. Analysis of recurrence pattern revealed neither a difference in the size of the recurrent tumor nor in the distance of the recurrences from the preoperative tumor location between groups. Overall frequencies of distant recurrences were 20 % in the reference group and 19 % (4/21 patients) in the cilengitide group. Compared with TMZ/RT → TMZ alone, the addition of cilengitide does not alter patterns of progression. This analysis does not support concerns that integrin antagonism by cilengitide may induce a more aggressive phenotype at progression, but also provides no evidence for an anti-invasive activity of cilengitide in patients with newly diagnosed glioblastoma.

Chondrocytes expressing intracellular collagen type II enter the cell cycle and co-express collagen type I in monolayer culture.

For autologous chondrocyte transplantation, articular chondrocytes are harvested from cartilage tissue and expanded in vitro in monolayer culture. We aimed to characterize with a cellular resolution the synthesis of collagen type II (COL2) and collagen type I (COL1) during expansion in order to further understand why these cells lose the potential to form cartilage tissue when re-introduced into a microenvironment that supports chondrogenesis. During expansion for six passages, levels of transcripts encoding COL2 decreased to <0.1%, whereas transcript levels encoding COL1 increased 370-fold as compared to primary chondrocytes. Flow cytometry for intracellular proteins revealed that chondrocytes acquired a COL2/COL1-double positive phenotype during expansion, and the COL2 positive cells were able to enter the cell cycle. While the fraction of COL2 positive cells decreased from 70% to <2% in primary chondrocytes to passage six cells, the fraction of COL1 positive cells increased from <1% to >95%. In parallel to the decrease of the fraction of COL2 positive cells, the cells' potential to form cartilage-like tissue in pellet cultures steadily decreased. Intracellular staining for COL2 enables for characterization of chondrocyte lineage cells in more detail with a cellular resolution, and it may allow predicting the effectiveness of expanded chondrocytes to form cartilage-like tissue.

Polyfunctional Two- (2D) and Three- (3D) Dimensional Oxalate Bridged Bimetallic Magnets

We report major results concerning polyfunctional two- (2D) and three- (3D) dimensional oxalate bridged bimetallic magnets. As a consequence of their specific organisation they are composed of an anionic sub-lattice and a cationic counter-part. These bimetallic polymers can accommodate various counter-cations possessing specific physical properties in addition to the magnetic ones resulting from the interactions between the metallic ions in the anionic sub-lattice. Thus, molecular magnets possessing paramagnetic, conductive and optical properties are presented in this review.

Cohort Profile: The Interdisciplinary Study of Inequalities in Smoking (ISIS)

The Interdisciplinary Study of Inequalities in Smoking (ISIS) is a cohort study investigating the joint effects of residents' socio-demographic characteristics and neighbourhood attributes on the social distribution of smoking in a young adult population. Smoking is a behaviour with an increasingly steep social class gradient; smoking prevalence among young adults is no longer declining at the same rate as among the rest of the population, and there is evidence of growing place-based disparities in smoking. ISIS was established to examine these pressing concerns. The ISIS sample comprises non-institutionalized individuals aged 18-25 years, who are proficient in English and/or French and who had been living at their current address in Montréal, Canada, for at least 1 year at time of first contact. Two waves of data have been collected: baseline data were collected November 2011-September 2012 (n = 2093), and a second wave of data was collected January-June 2014 (n = 1457). Data were collected from respondents using a self-administered questionnaire, developed by the research team based on sociological theory, which includes questions concerning social, economic, cultural and biological capital, and activity space as well as smoking behaviour. Data are available upon request from [katherine.frohlich@umontreal.ca].