Use Of Experimental Resources Research Articles

Enhanced Sequence-Activity Mapping and Evolution of Artificial Metalloenzymes by Active Learning.

Tailored enzymes are crucial for the transition to a sustainable bioeconomy. However, enzyme engineering is laborious and failure-prone due to its reliance on serendipity. The efficiency and success rates of engineering campaigns may be improved by applying machine learning to map the sequence-activity landscape based on small experimental data sets. Yet, it often proves challenging to reliably model large sequence spaces while keeping the experimental effort tractable. To address this challenge, we present an integrated pipeline combining large-scale screening with active machine learning, which we applied to engineer an artificial metalloenzyme (ArM) catalyzing a new-to-nature hydroamination reaction. Combining lab automation and next-generation sequencing, we acquired sequence-activity data for several thousand ArM variants. We then used Gaussian process regression to model the activity landscape and guide further screening rounds. Critical characteristics of our pipeline include the cost-effective generation of information-rich data sets, the integration of an explorative round to improve the model's performance, and the inclusion of experimental noise. Our approach led to an order-of-magnitude boost in the hit rate while making efficient use of experimental resources. Search strategies like this should find broad utility in enzyme engineering and accelerate the development of novel biocatalysts.

Evolutionary game analysis of stakeholders’ decision-making behavior in agricultural data supply chain

The significance of agricultural information sharing in fostering agricultural development cannot be overstated. This practice plays a pivotal role in disseminating cutting-edge agricultural technologies, cultivation methods, and pest control strategies, empowering farmers with valuable knowledge to enhance crop yield and quality. Moreover, it aligns with government objectives of resource sharing and addressing gaps, contributing to the advancement of agricultural modernization and the development of the industry chain. Despite its inherent benefits, the practical implementation of agricultural information sharing faces challenges. Stakeholders engaged in information sharing often prioritize individual benefits, potentially leading to a decline in agricultural information quality and the inefficient use of experimental resources. To confront this issue, the present research establishes a three-party evolutionary game model comprising an agricultural product data sharing platform, agricultural data providers, and agricultural data consumers. Leveraging dynamic system theory, the model analyzes the evolutionary stable strategies of stakeholders and investigates the critical factors influencing the strategic choices of these three parties. Experimental findings underscore the pivotal role of participants’ initial strategies, regulatory intensity, reward and punishment mechanisms, and information feedback in shaping stakeholder decision-making behavior. Implementation of measures such as heightened scrutiny of information on the sharing platform and fostering consumer trust in data emerges as imperative for enhancing system stability. These actions are essential for constructing an efficient and reliable information-sharing ecosystem, thereby facilitating the sustainable development of modern agriculture.

Design Amortization for Bayesian Optimal Experimental Design

Bayesian optimal experimental design is a sub-field of statistics focused on developing methods to make efficient use of experimental resources. Any potential design is evaluated in terms of a utility function, such as the (theoretically well-justified) expected information gain (EIG); unfortunately however, under most circumstances the EIG is intractable to evaluate. In this work we build off of successful variational approaches, which optimize a parameterized variational model with respect to bounds on the EIG. Past work focused on learning a new variational model from scratch for each new design considered. Here we present a novel neural architecture that allows experimenters to optimize a single variational model that can estimate the EIG for potentially infinitely many designs. To further improve computational efficiency, we also propose to train the variational model on a significantly cheaper-to-evaluate lower bound, and show empirically that the resulting model provides an excellent guide for more accurate, but expensive to evaluate bounds on the EIG. We demonstrate the effectiveness of our technique on generalized linear models, a class of statistical models that is widely used in the analysis of controlled experiments. Experiments show that our method is able to greatly improve accuracy over existing approximation strategies, and achieve these results with far better sample efficiency.

Quantum simulation of lattice gauge theories using Wilson fermions

Quantum simulators have the exciting prospect of giving access to real-time dynamics of lattice gauge theories, in particular in regimes that are difficult to compute on classical computers. Future progress towards scalable quantum simulation of lattice gauge theories, however, hinges crucially on the efficient use of experimental resources. As we argue in this work, due to the fundamental non-uniqueness of discretizing the relativistic Dirac Hamiltonian, the lattice representation of gauge theories allows for an optimization that up to now has been left unexplored. We exemplify our discussion with lattice quantum electrodynamics in two-dimensional space-time, where we show that the formulation through Wilson fermions provides several advantages over the previously considered staggered fermions. Notably, it enables a strongly simplified optical-lattice setup and it reduces the number of degrees of freedom required to simulate dynamical gauge fields. Exploiting the optimal representation, we propose an experiment based on a mixture of ultracold atoms trapped in a tilted optical lattice. Using numerical benchmark simulations, we demonstrate that a state-of-the-art quantum simulator may access the Schwinger mechanism and map out its non-perturbative onset.

Automatic Between-Pulse Analysis of DIII-D Experimental Data Performed Remotely on a Supercomputer at Argonne Leadership Computing Facility

For the first time, an automatically triggered, between-pulse fusion science analysis code was run on-demand at a remotely located supercomputer at Argonne Leadership Computing Facility (ALCF, Lemont, Illinois) in support of in-process experiments being performed at DIII-D (San Diego, California). This represents a new paradigm for combining geographically distant experimental and high-performance computing facilities to provide enhanced data analysis that is quickly available to researchers. Enhanced analysis improves the understanding of the current pulse, translating into a more efficient use of experimental resources and quality of the resultant science. The analysis code used here, called SURFMN, calculates the magnetic structure of the plasma using Fourier transform. Increasing the number of Fourier components provides a more accurate determination of the stochastic boundary layer near the plasma edge by better resolving magnetic islands, but requires 26 min to complete using local DIII-D resources, putting it well outside the useful time range for between-pulse analysis. These islands relate to confinement and edge-localized mode suppression, and may be controlled by adjusting coil currents for the next pulse. ALCF has ensured on-demand execution of SURFMN by providing a reserved queue, a specialized service that launches the code after receiving an automatic trigger, and network access from the worker nodes for data transfer. Runs are executed on 252 cores of ALCF’s Cooley cluster and the data are available locally at DIII-D within 3 min of triggering. The original SURFMN design limits additional improvements with more cores; however, our work shows a path forward where codes that benefit from thousands of processors can run between pulses.

Learning from complementary ways of developing experimental competences

Engineering education has solid needs of experimental competences development. Nowadays theses competences can be developed not only in traditional laboratories (hands on) but also through the use of computer simulations and remote labs. The use of diversified methods in education and the exploration of new resources and techniques in classroom may allow teachers to motivate more students, and capture their attention due to their different learning styles.The main objective of this thesis project is to better understand the effects on the students’ learning outcomes in different contexts, when subject to similar design approaches using an enquiry-based teaching and learning methodology based on simultaneous use of experimental resources (hands on, simulation and remote labs) together with calculus, in class and assessment. To accomplish this goal, several insights must be taken into consideration, including the teachers’ mediation in each case and the didactical implementations adaptations, but also external factors, such as socio-cultural and/or political factors.

Rationally reduced libraries for combinatorial pathway optimization minimizing experimental effort.

Rational flux design in metabolic engineering approaches remains difficult since important pathway information is frequently not available. Therefore empirical methods are applied that randomly change absolute and relative pathway enzyme levels and subsequently screen for variants with improved performance. However, screening is often limited on the analytical side, generating a strong incentive to construct small but smart libraries. Here we introduce RedLibs (Reduced Libraries), an algorithm that allows for the rational design of smart combinatorial libraries for pathway optimization thereby minimizing the use of experimental resources. We demonstrate the utility of RedLibs for the design of ribosome-binding site libraries by in silico and in vivo screening with fluorescent proteins and perform a simple two-step optimization of the product selectivity in the branched multistep pathway for violacein biosynthesis, indicating a general applicability for the algorithm and the proposed heuristics. We expect that RedLibs will substantially simplify the refactoring of synthetic metabolic pathways.

A practical guide for optimal designs of experiments in the Monod model

The Monod model is a classical microbiological model often used in environmental sciences, for example to evaluate biodegradation processes. The model describes microbial growth kinetics in batch culture experiments using three parameters: the maximal specific growth rate, the saturation constant and the yield coefficient. However, identification of these parameter values from experimental data is a challenging problem. Recently, it was demonstrated theoretically that the application of optimal design theory in this model is an efficient method for both parameter value identification and economic use of experimental resources [Dette, H., Melas, V.B., Pepelyshev, A., Strigul, N., 2003. Efficient design of experiments in the Monod model. J. R. Stat. Soc. B 65 (Part 3), 725–742]. The purpose of this paper is to provide this method as a computational “tool” such that it can be used by practitioners without a strong mathematical and statistical background for the efficient design of experiments in the Monod model. The paper presents careful explanations of the principal theoretical concepts, and a simple algorithm for practical optimal design calculations.

Development and optimization of NO x storage and reduction catalysts using statistically guided high-throughput experimentation

The interaction between reaction conditions and catalyst compositions for NO x storage and reduction (NSR) catalysts has been explored using a combination of high-throughput experimentation and experimental design. This was accomplished by varying the concentration of NO and O 2, the reducing agent concentration and identity, space velocity, and reactor temperature. All reaction conditions were varied simultaneously for a range of Pt, Ba, and Fe loadings on γ-Al 2O 3. It was found that most of the reaction conditions investigated affected the saturation NO x storage, the production of nitrous oxide, and the steady state lean NO x reduction. An empirical model was developed using response surface methodology to predict the saturation NO x storage and nitrous oxide production as a function of Pt, Ba, and Fe weight loading. This model was used to optimize the catalytic composition and a comparison between the model predictions and experimental results for the optimized catalysts are given. It was shown that high-throughput experimentation in combination with experimental design leads to more efficient use of experimental resources in addition to a more in depth understanding of catalytic systems.

Combinatorial Materials Development using Gradient Arrays. Sphere Covering Lattice Designs: Designs for Efficient use of Experimental Resources.

ABSTRACTGradient arrays are now common tools in combinatorial chemistry for discovery of new leads to commercial materials. Although the cost per sample has dropped markedly with new high throughput methods, efficient use of experimental resources is still important. Examination of gradient arrays from an informational standpoint suggests that designs which use the concepts of sphere packing and covering will be more efficient than simple gradients. This is especially true in higher dimensional systems.

Publishing, interpreting, and accessing models

Models are tools that are used to predict the structure and the behavior of a system. They are usually developed to fit a particular set of data, however, once developed, they can be applied to other studies; to predict the behavior of the system under different conditions, explore the structure of parts of the system that were not sampled, identify areas that are lacking in data, and optimize the design of new studies. For published models to be interpreted and applied widely to other studies, models need to be communicated clearly. Papers describing models should contain full descriptions of the physiological relevance and mathematical rigor of the model. A working version of the model then needs to be made available to other investigators. This manuscript examines the areas of communicating and interpreting published models and describes a facility, called a ‘model library’, that is being developed to make published models available on the Internet. With better communication, information in models can be extracted and used to bridge the gap between theory and experiment to extend our understanding of complex systems and optimize the use of experimental resources.