Publications

Journal Article Double stranded DNA sequencing as a choice for DNA sequencing Get access Hong Zhang, Hong Zhang Search for other works by this author on: Oxford Academic PubMed Google Scholar Randy Scholl, Randy Scholl 1Genetics Department, Ohio State UniversityColumbus, OH 43210, USA Search for other works by this author on: Oxford Academic PubMed Google Scholar John Browse, John Browse Search for other works by this author on: Oxford Academic PubMed Google Scholar Chris Somerville Chris Somerville Search for other works by this author on: Oxford Academic PubMed Google Scholar Nucleic Acids Research, Volume 16, Issue 3, 11 February 1988, Page 1220, https://doi.org/10.1093/nar/16.3.1220 Published: 11 February 1988 Article history Accepted: 23 October 1987 Published: 11 February 1988

A micellar assembly of molecules constituted of poly(ethylene glycol) as the hydrophilic component and 2-diazo-1,2-naphthoquinone as the hydrophobic component was shown to be destroyed in a two-photon photoreaction triggered by infrared light with release of an encapsulated fluorescent probe molecule.

An entry from the Inorganic Crystal Structure Database, the world’s repository for inorganic crystal structures. The entry contains experimental data from a crystal diffraction study. The deposited dataset for this entry is freely available from the joint CCDC and FIZ Karlsruhe Access Structures service and typically includes 3D coordinates, cell parameters, space group, experimental conditions and quality measures.

Abstract Although chemical and enzymatic catalysts have been combined, reactions in which an organometallic catalyst and a metalloenzyme work cooperatively to create products, which cannot be generated with either catalyst alone or in comparable yields by sequential reactions of the two catalysts, have not been reported. Such reactions are challenging to achieve, in part because the milieu in which these catalysts operate are typically different. Herein, two classes of catalysts are demonstrated to react cooperatively in the same system. Combination of a metathesis catalyst and a P450 enzyme lead to a dynamic equilibration of alkenes and a selective epoxidation of the cross‐metathesis products. These results show the potential of combining the two classes of catalysts for synthetic transformations.

Feature stores (also sometimes referred to as embedding stores) are becoming ubiquitous in model serving systems: downstream applications query these stores for auxiliary inputs at inference-time. Stored features are derived by featurizing rapidly changing base data sources. Featurization can be costly prohibitively expensive to trigger on every data update, particularly for features that are vector embeddings computed by a model. Yet, existing systems naively apply a one-size-fits-all policy as to when/how to update these features, and do not consider query access patterns or impacts on prediction accuracy. This paper introduces RALF, which orchestrates feature updates by leveraging downstream error feedback to minimize feature store regret , a metric for how much featurization degrades downstream accuracy. We evaluate with representative feature store workloads, anomaly detection and recommendation, using real-world datasets. We run system experiments with a 275,077 key anomaly detection workload on 800 cores to show up to a 32.7% reduction in prediction error or up to 1.6X compute cost reduction with accuracy-aware scheduling.

Query optimizers are a performance-critical component in every database system. Due to their complexity, optimizers take experts months to write and years to refine. In this work, we demonstrate for the first time that learning to optimize queries without learning from an expert optimizer is both possible and efficient. We present Balsa, a query optimizer built by deep reinforcement learning. Balsa first learns basic knowledge from a simple, environment-agnostic simulator, followed by safe learning in real execution. On the Join Order Benchmark, Balsa matches the performance of two expert query optimizers, both open-source and commercial, with two hours of learning, and outperforms them by up to 2.8$\times$ in workload runtime after a few more hours. Balsa thus opens the possibility of automatically learning to optimize in future compute environments where expert-designed optimizers do not exist.

The paper reports on an investigation of the fatigue failure of tree retardant cross-linked polyethylene (TR-XLPE) that is relevant to water tree development in underground cable insulation. Finite element calculations were used to estimate the stresses developed in cable insulation by di-electrophoretic forces; these stresses are in the low megaPascal range around inclusions (or water tree branches) that are long and thin. They are insufficient to bring about instantaneous failure of the insulation. However, these stresses might be sufficient to cause cyclic fatigue failure of the insulation, and, accordingly, fatigue measurements were carried out on samples from a commercial cable. The resulting fatigue failures, that occurred at cycle numbers achievable in practical work, suggest that fatigue might be an explanation for slow development of water trees over years of service in the ground. Cycle numbers at failure were found to be lower at higher mechanical stresses, at higher temperatures and in the presence of humic acid or ferric ions; however, the number of cycles to failure was larger in the presence of water.

We present a large-scale density functional theory (DFT) investigation of the ABO(3) chemical space in the perovskite crystal structure, with the aim of identifying those that are relevant for forming piezoelectric materials. Screening criteria on the DFT results are used to select 49 compositions, which can be seen as the fundamental building blocks from which to create alloys with potentially good piezoelectric performance. This screening finds all the alloy end points used in three well-known high-performance piezoelectrics. The energy differences between different structural distortions, deformation, coupling between the displacement of the A and B sites, spontaneous polarization, Born effective charges, and stability is analyzed in each composition. We discuss the features that cause the high piezoelectric performance of the well-known piezoelectric lead zirconate titanate (PZT), and investigate to what extent these features occur in other compositions. We demonstrate how our results can be useful in the design of isovalent alloys with high piezoelectric performance.

The tunable physical and electronic structure of III-V semiconductor alloys renders them uniquely useful for a variety of applications, including biological imaging, transistors, and solar energy conversion. However, their fabrication typically requires complex gas phase instrumentation or growth from high-temperature melts, which consequently limits their prospects for widespread implementation. Furthermore, the need for lattice matched growth substrates in many cases confines the composition of the materials to a narrow range that can be epitaxially grown. In this work, we present a solution phase synthesis for indium gallium phosphide (In(x)Ga(1-x)P) alloy nanowires, whose indium/gallium ratio, and consequently, physical and electronic structure, can be tuned across the entire x = 0 to x = 1 composition range. We demonstrate the evolution of structural and optical properties of the nanowires, notably the direct to indirect band gap transition, as the composition is varied from InP to GaP. Our scalable, low-temperature synthesis affords compositional, structural, and electronic tunability and can provide a route for realization of broader In(x)Ga(1-x)P applications.