Understanding climate phenomena with data-driven models

In climate science, climate models are one of the main tools for understanding phenomena. Here, we develop a framework to assess the fitness of a climate model for providing understanding. The framework is based on three dimensions: representational accuracy, representational depth, and graspability. We show that this framework does justice to the intuition that classical process-based climate models give understanding of phenomena. While simple climate models are characterized by a larger graspability, state-of-the-art models have a higher representational accuracy and representational depth. We then compare the fitness-for-providing understanding of process-based to data-driven models that are built with machine learning. We show that at first glance, data-driven models seem either unnecessary or inadequate for understanding. However, a case study from atmospheric research demonstrates that this is a false dilemma. Data-driven models can be useful tools for understanding, specifically for phenomena for which scientists can argue from the coherence of the models with background knowledge to their representational accuracy and for which the model complexity can be reduced such that they are graspable to a satisfactory extent.     

Genomic and metabolic prediction of complex heterotic traits in hybrid maize

Maize is both an exciting model organism in plant genetics and also the most important crop worldwide for food, animal feed and bioenergy production. Recent genome-wide association and metabolic profiling studies aimed to resolve quantitative traits to their causal genetic loci and key metabolic regulators. Here we present a complementary approach that exploits large-scale genomic and metabolic information to predict complex, highly polygenic traits in hybrid testcrosses. We crossed 285 diverse Dent inbred lines from worldwide sources with two testers and predicted their combining abilities for seven biomass- and bioenergy-related traits using 56,110 SNPs and 130 metabolites. Whole-genome and metabolic prediction models were built by fitting effects for all SNPs or metabolites. Prediction accuracies ranged from 0.72 to 0.81 for SNPs and from 0.60 to 0.80 for metabolites, allowing a reliable screening of large collections of diverse inbred lines for their potential to create superior hybrids.     

Chromosomal rearrangements maintain a polymorphic supergene controlling butterfly mimicry

Supergenes are tight clusters of loci that facilitate the co-segregation of adaptive variation, providing integrated control of complex adaptive phenotypes. Polymorphic supergenes, in which specific combinations of traits are maintained within a single population, were first described for 'pin' and 'thrum' floral types in Primula and Fagopyrum, but classic examples are also found in insect mimicry and snail morphology. Understanding the evolutionary mechanisms that generate these co-adapted gene sets, as well as the mode of limiting the production of unfit recombinant forms, remains a substantial challenge. Here we show that individual wing-pattern morphs in the polymorphic mimetic butterfly Heliconius numata are associated with different genomic rearrangements at the supergene locus P. These rearrangements tighten the genetic linkage between at least two colour-pattern loci that are known to recombine in closely related species, with complete suppression of recombination being observed in experimental crosses across a 400-kilobase interval containing at least 18 genes. In natural populations, notable patterns of linkage disequilibrium (LD) are observed across the entire P region. The resulting divergent haplotype clades and inversion breakpoints are found in complete association with wing-pattern morphs. Our results indicate that allelic combinations at known wing-patterning loci have become locked together in a polymorphic rearrangement at the P locus, forming a supergene that acts as a simple switch between complex adaptive phenotypes found in sympatry. These findings highlight how genomic rearrangements can have a central role in the coexistence of adaptive phenotypes involving several genes acting in concert, by locally limiting recombination and gene flow.     

Virtual cranial reconstruction of Sahelanthropus tchadensis

Previous research in Chad at the Toros-Menalla 266 fossiliferous locality (about 7 million years old) uncovered a nearly complete cranium (TM 266-01-60-1), three mandibular fragments and several isolated teeth attributed to Sahelanthropus tchadensis. Of this material, the cranium is especially important for testing hypotheses about the systematics and behavioural characteristics of this species, but is partly distorted from fracturing, displacement and plastic deformation. Here we present a detailed virtual reconstruction of the TM 266 cranium that corrects these distortions. The reconstruction confirms that S. tchadensis is a hominid and is not more closely related to the African great apes. Analysis of the basicranium further indicates that S. tchadensis might have been an upright biped, suggesting that bipedalism was present in the earliest known hominids, and probably arose soon after the divergence of the chimpanzee and human lineages.     

Aneuploidy and cancer

In contrast to normal cells, aneuploidy--alterations in the number of chromosomes--is consistently observed in virtually all cancers. A growing body of evidence suggests that aneuploidy is often caused by a particular type of genetic instability, called chromosomal instability, which may reflect defects in mitotic segregation in cancer cells. A better understanding of the molecular mechanisms leading to aneuploidy holds promise for the development of cancer drugs that target this process.     

Evolutionary dynamics on graphs

Evolutionary dynamics have been traditionally studied in the context of homogeneous or spatially extended populations. Here we generalize population structure by arranging individuals on a graph. Each vertex represents an individual. The weighted edges denote reproductive rates which govern how often individuals place offspring into adjacent vertices. The homogeneous population, described by the Moran process, is the special case of a fully connected graph with evenly weighted edges. Spatial structures are described by graphs where vertices are connected with their nearest neighbours. We also explore evolution on random and scale-free networks. We determine the fixation probability of mutants, and characterize those graphs for which fixation behaviour is identical to that of a homogeneous population. Furthermore, some graphs act as suppressors and others as amplifiers of selection. It is even possible to find graphs that guarantee the fixation of any advantageous mutant. We also study frequency-dependent selection and show that the outcome of evolutionary games can depend entirely on the structure of the underlying graph. Evolutionary graph theory has many fascinating applications ranging from ecology to multi-cellular organization and economics.     

Oncogene-induced senescence as an initial barrier in lymphoma development

Acute induction of oncogenic Ras provokes cellular senescence involving the retinoblastoma (Rb) pathway, but the tumour suppressive potential of senescence in vivo remains elusive. Recently, Rb-mediated silencing of growth-promoting genes by heterochromatin formation associated with methylation of histone H3 lysine 9 (H3K9me) was identified as a critical feature of cellular senescence, which may depend on the histone methyltransferase Suv39h1. Here we show that Emicro-N-Ras transgenic mice harbouring targeted heterozygous lesions at the Suv39h1, or the p53 locus for comparison, succumb to invasive T-cell lymphomas that lack expression of Suv39h1 or p53, respectively. By contrast, most N-Ras-transgenic wild-type ('control') animals develop a non-lymphoid neoplasia significantly later. Proliferation of primary lymphocytes is directly stalled by a Suv39h1-dependent, H3K9me-related senescent growth arrest in response to oncogenic Ras, thereby cancelling lymphomagenesis at an initial step. Suv39h1-deficient lymphoma cells grow rapidly but, unlike p53-deficient cells, remain highly susceptible to adriamycin-induced apoptosis. In contrast, only control, but not Suv39h1-deficient or p53-deficient, lymphomas senesce after drug therapy when apoptosis is blocked. These results identify H3K9me-mediated senescence as a novel Suv39h1-dependent tumour suppressor mechanism whose inactivation permits the formation of aggressive but apoptosis-competent lymphomas in response to oncogenic Ras.