Dataset Information

Using Differential Evolution to Design Optimal Experiments.

ABSTRACT: Differential Evolution (DE) has become one of the leading metaheuristics in the class of Evolutionary Algorithms, which consists of methods that operate off of survival-of-the-fittest principles. This general purpose optimization algorithm is viewed as an improvement over Genetic Algorithms, which are widely used to find solutions to chemometric problems. Using straightforward vector operations and random draws, DE can provide fast, efficient optimization of any real, vector-valued function. This article reviews the basic algorithm and a few of its modifications with various enhancements. We provide guidance for practitioners, discuss implementation issues and give illustrative applications of DE with the corresponding R codes to find different types of optimal designs for various statistical models in chemometrics that involve the Arrhenius equation, reaction rates, concentration measures and chemical mixtures.

SUBMITTER: Stokes Z

PROVIDER: S-EPMC7088454 | biostudies-literature | 2020 Apr

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Using Differential Evolution to Design Optimal Experiments.

Stokes Zack Z Mandal Abhyuday A Wong Weng Kee WK

Chemometrics and intelligent laboratory systems : an international journal sponsored by the Chemometrics Society 20200128

Differential Evolution (DE) has become one of the leading metaheuristics in the class of Evolutionary Algorithms, which consists of methods that operate off of survival-of-the-fittest principles. This general purpose optimization algorithm is viewed as an improvement over Genetic Algorithms, which are widely used to find solutions to chemometric problems. Using straightforward vector operations and random draws, DE can provide fast, efficient optimization of any real, vector-valued function. Thi ...[more]

PMID: 32205900

Similar Datasets

Project description:MotivationWe addressed the problem of inferring gene regulatory network (GRN) from gene expression data of knockout (KO) experiments. This inference is known to be underdetermined and the GRN is not identifiable from data. Past studies have shown that suboptimal design of experiments (DOE) contributes significantly to the identifiability issue of biological networks, including GRNs. However, optimizing DOE has received much less attention than developing methods for GRN inference.ResultsWe developed REDuction of UnCertain Edges (REDUCE) algorithm for finding the optimal gene KO experiment for inferring directed graphs (digraphs) of GRNs. REDUCE employed ensemble inference to define uncertain gene interactions that could not be verified by prior data. The optimal experiment corresponds to the maximum number of uncertain interactions that could be verified by the resulting data. For this purpose, we introduced the concept of edge separatoid which gave a list of nodes (genes) that upon their removal would allow the verification of a particular gene interaction. Finally, we proposed a procedure that iterates over performing KO experiments, ensemble update and optimal DOE. The case studies including the inference of Escherichia coli GRN and DREAM 4 100-gene GRNs, demonstrated the efficacy of the iterative GRN inference. In comparison to systematic KOs, REDUCE could provide much higher information return per gene KO experiment and consequently more accurate GRN estimates.ConclusionsREDUCE represents an enabling tool for tackling the underdetermined GRN inference. Along with advances in gene deletion and automation technology, the iterative procedure brings an efficient and fully automated GRN inference closer to reality.Availability and implementationMATLAB and Python scripts of REDUCE are available on www.cabsel.ethz.ch/tools/REDUCE CONTACT: rudi.gunawan@chem.ethz.chSupplementary informationSupplementary data are available at Bioinformatics online.

Dataset Information

Using Differential Evolution to Design Optimal Experiments.

Publications

Using Differential Evolution to Design Optimal Experiments.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets