Editors' Choice

Science  01 Apr 2016:
Vol. 352, Issue 6281, pp. 49
  1. Galaxy Simulations

    Machine learning in cosmological models

    1. Keith T. Smith

    Machine learning can help model the formation and evolution of galaxies


    A cosmological simulation containing only dark matter is relatively easy to run, but adding gas, stars, and galaxies to the model requires sophisticated hydrodynamics and subgrid physics. These are very computationally expensive, but Kamdar et al. have used a machine learning algorithm to massively speed up the process. By training the algorithm on part of a large hydrodynamic simulation, they are able to reproduce many galactic properties in the rest of the simulation just from the dark matter information, and in much less time. The technique could be used to quickly scale up detailed hydrodynamic simulations to larger dark matter–only ones, aiding in the interpretation of observational surveys.

    Mon. Not. R. Astron. Soc. 457, 1162 (2016).

  2. Germline

    Soma regulates germline differentiation

    1. Beverly A. Purnell

    In the Drosophila ovary, germline stem cells (GSCs) generate a cystoblast containing 16 cells, only one of which becomes an oocyte. Somatic cells adjacent to germ cells influence cystoblast proliferation and differentiation. Upadhyay et al. show that the histone methyltransferase dSETDB1, located in escort cells that nurture the cystoblast, promotes differentiation and controls GSCs through the signaling factor dWnt4, which in turn regulates adherens junction proteins in escort cells to promote cystoblast encapsulation and differentiation. Parallel to this pathway, transposable elements also regulate dWnt4 for CB encapsulation and differentiation. Hence, germ cells are influenced by surrounding somatic cells, and this work links the somatically expressed histone modifier and transposon to Wnt molecule activity for cell differentiation.

    PLOS.Genet. 10.1371/journal.pgen.1005918 (2016).

  3. Tetrapod Evolution

    Water walker

    1. Sacha Vignieri

    The ancestor of all tetrapods “walked” out of the water millions of years ago. An intriguing idea, but walking out of water is actually quite a biomechanical challenge. Though some fishes are known to use their fins to “walk” underwater, the movement toward a pelvic girdle that could support the weight of an organism's body out of the water has been seen as a tetrapod innovation. Flammang et al. analyzed the pelvic structure and walking kinematics of a rare cave fish, Cryptotora thamicola, which is known to walk up waterfalls using its fins and a tetrapod-like lateral gait, and found remarkable convergence with the tetrapod pelvis. This finding supports the hypothesis that pelvically driven movement on land may have been possible before the evolution of digited limbs.

    Sci. Rep. 10.1038/srep23711 (2016).

    The waterfall-climbing cave fish uses a tetrapod-like gait to climb waterfalls

  4. Aging

    Reversing vascular deterioration in aged mice

    1. L. Bryan Ray

    Stiff old arteries contribute to cardiovascular disease in the elderly, which is a leading cause of death. De Picciotto et al. report that age-related deterioration in the flexibility of the carotid artery in mice could be reversed when animals received dietary supplementation of nicotinamide mononucleotide (NMN). NMN is an intermediate in the synthesis of NAD+ (the reduced form of nicontinamide adenine dinucleotide), which improved metabolic function and stress responses in older animals. Treatment of mice with NMN for 8 weeks improved measures of elasticity in large arteries. NMN may act, at least in part, by activating sirtuin 1, an NAD+-dependent protein deacetylase. Dietary supplementation of NMN may thus provide a therapeutic strategy to reverse arterial dysfunction in the elderly.

    Aging Cell 10.1111/acel.12461 (2016).

  5. Diversity in Science

    Mismatch reduces minority STEM success

    1. Brad Wible

    A policy aimed at increasing the enrollment of minority undergraduates in STEM (science, technology, engineering, and mathematics) majors actually hurt overall minority representation in STEM. Arcidiacono et al. studied all students who enrolled at any University of California campus from 1995 to 1997, when race was a factor in admissions. Many minority students were admitted to STEM majors at the most competitive campuses, such as Berkeley and Los Angeles, despite high-school grades and test scores that were better aligned with those of incoming students at less competitive campuses, such as Santa Cruz and Riverside. Switches to non-STEM majors and failures to graduate, particularly among minorities that were underprepared for the most competitive campuses, could have been minimized had school admissions been better matched to precollege preparation, boosting overall minority representation among STEM graduates.

    Am. Econ. Rev. 106, 525 (2016).

  6. Polymers

    Tough polymers from renewable sources

    1. Phil Szuromi

    Three renewable sources have been combined to create a tough thermoplastic. Zhang et al. used hydroxypropyl methylcellulose (HPMC) as a semirigid backbone and grafted on rubbery poly(β-methyl-δ-valerolactone) (PMVL) blocks that were then terminated with hard poly-l-lactide (PLLA) blocks. The rubbery block allowed the PLLA blocks to crystallize. These materials were easily processed in the molten state and showed exceptional toughness: They could be strained up to 600% and yet remained transparent up to the point of pulling apart.

    ACS Macro Lett. 5, 407 (2016).

  7. Disease Ecology

    Forecasting forest recovery

    1. Andrew M. Sugden

    Models can help predict what species will replace ash trees lost to disease


    Fungal pathogens of trees, particularly of common species, can have pronounced effects on the composition and functioning of the wider forest community. Needham et al. develop a predictive model of the effects of ash dieback, a disease that has recently spread across Europe, on the population dynamics and community structure of the remaining components of a forest community in the United Kingdom. They use demographic data on the current forest community to forecast which species are likely to replace ash (Fraxinus excelsior) after the disease strikes, as it is expected to do within the next few years. The likely beneficiary in this case is sycamore (Acer pseudoplatanus), which is expected to be in the best position to exploit the resources and space left by the dieback of ash. The modeling approach could be deployed to help predict the course of recovery after other tree diseases, in forests where inventories and demographic time series data have been assembled.

    J. Ecol. 104, 315 (2016).

Navigate This Article