Computational and statistical tradeoffs via convex relaxation
Venkat Chandrasekaran and Michael I. Jordan
The growth in the size and scope of datasets in science and technology has created a need for foundational perspectives on data analysis that blend computer science and statistics. Specifically, the core challenge with massive datasets is that of guaranteeing improved accuracy of an analysis procedure as data accrue, even in the face of a time budget. We address this problem (pp. E1181–E1190) via a notion of “algorithmic weakening,” whereby as data scale, the procedure backs off to cheaper algorithms, leveraging the growing inferential strength of the data to ensure that a desired level of accuracy is achieved within the computational budget.
Prediction of function for the polyprenyl transferase subgroup in the isoprenoid synthase superfamily
Frank H. Wallrapp, Jian-Jung Pan, Gurusankar Ramamoorthy, Daniel E. Almonacid, Brandan S. Hillerich, Ronald Seidel, Yury Patskovsky, Patricia C. Babbitt, Steven C. Almo, Matthew P. Jacobson, and C. Dale Poulter
This paper (pp. E1196–E1202) reports a large-scale collaborative study of an approach for predicting the function of chain elongation prenyltransferases from genetic data. A diverse set of genes for enzymes in the isoprenoid synthase superfamily was identified for cloning, expression, X-ray structural analysis, and prediction of function by docking to homology models. Blind predictions, later verified biochemically, were accurate to within one isoprene unit for all but a few of the 74 enzymes studied, an extraordinarily high level of prediction given that the enzymes often give products whose chain lengths vary by one isoprene unit.
Steroid-based facial amphiphiles for stabilization and crystallization of membrane proteins
Sung Chang Lee, Brad C. Bennett, Wen-Xu Hong, Yu Fu, Kent A. Baker, Julien Marcoux, Carol V. Robinson, Andrew B. Ward, James R. Halpert, Raymond C. Stevens, Charles David Stout, Mark J. Yeager, and Qinghai Zhang
Membrane proteins (MPs) perform a variety of essential cellular functions, account for about one-third of encoded proteins in genomes, and comprise more than one-half of human drug targets. High-resolution structures are essential to understand the underlying molecular mechanisms of MPs and facilitate structure-based drug design efforts. Detergents are indispensible in the solubilization of MPs, but they tend to destabilize MPs and often impede the growth of well-ordered protein crystals. We describe (pp. E1203–E1211) a class of structurally unique detergents, designated as facial amphiphiles, which improved MP stability and success in the crystallization of different families of MPs.
Complete and unidirectional conversion of human embryonic stem cells to trophoblast by BMP4
Mitsuyoshi Amita, Katsuyuki Adachi, Andrei P. Alexenko, Sunilima Sinha, Danny J. Schust, Laura C. Schulz, R. Michael Roberts, and Toshihiko Ezashi
Human embryonic stem cells (hESC) exposed to the growth factor bone morphogenic protein 4 (BMP4) in the absence of FGF2 have been used to study the development of placental trophoblasts, but the soundness of this model has been challenged by others who concluded that the directional differentiation was primarily toward the mesoderm lineage rather than trophoblast. Here (pp. E1212–E1221) we identify key culture conditions necessary for BMP4 to convert hESC to an epithelium that expresses a full range of trophoblast markers, demonstrates invasive properties, and releases large quantities of placental hormones, with no evidence for mesoderm formation.
Binding of Drosophila Polo kinase to its regulator Matrimony is noncanonical and involves two separate functional domains
Amanda M. Bonner, Stacie E. Hughes, Jennifer A. Chisholm, S. Kendall Smith, Brian D. Slaughter, Jay R. Unruh, Kimberly A. Collins, Jennifer M. Friederichs, Laurence Florens, Selene K. Swanson, Marissa C. Pelot, Danny E. Miller, Michael P. Washburn, Sue L. Jaspersen, and R. Scott Hawley
Polo kinase regulates many processes during cell division and is upregulated in many cancers. During Drosophila female meiosis, the protein Matrimony inhibits Polo kinase using a noncanonical mechanism of Polo binding. Complete loss of Matrimony leads to meiotic catastrophe, and partial loss leads to chromosome missegregation. Proper Matrimony–Polo binding is required to prevent these defects, indicating that preventing Polo from phosphorylating targets is necessary for proper completion of meiosis. This finding is in contrast to mitosis where phosphorylation by Polo is usually required for cell division. This work (pp. E1222–E1231) provides important insight into developing anticancer therapeutic agents targeting Polo kinase.
High-resolution metabolic mapping of cell types in plant roots
Arieh Moussaieff, Ilana Rogachev, Leonid Brodsky, Sergey Malitsky, Ted W. Toal, Heather Belcher, Merav Yativ, Siobhan M. Brady, Philip N. Benfey, and Asaph Aharoni
Analyzing metabolite composition offers a powerful tool for understanding gene function and regulatory processes. Here (pp. E1232–E1241), we present nontargeted metabolomics assays of five Arabidopsis GFP-tagged lines representing core cell types in the plant root, providing a metabolic map of an organ, composed of its different cell types. Fifty metabolites were putatively identified. The most prominent groups were glucosinolates, phenylpropanoids, and dipeptides. Metabolites were differentially abundant across root cell types and in many cases, this abundance did not correlate with transcript expression, suggesting non–cell-autonomous mechanisms responsible for their targeted localization.