Multilevel modelling for inference of genetic regulatory networks

Time-course experiments with microarrays are often used to study dynamic biological systems and genetic regulatory networks (GRNs) that model how genes influence each other in cell-level development of organisms. The inference for GRNs provides important insights into the fundamental biological processes such as growth and is useful in disease diagnosis and genomic drug design. Due to the experimental design, multilevel data hierarchies are often present in time-course gene expression data. Most existing methods, however, ignore the dependency of the expression measurements over time and the correlation among gene expression profiles. Such independence assumptions violate regulatory interactions and can result in overlooking certain important subject effects and lead to spurious inference for regulatory networks or mechanisms. In this paper, a multilevel mixed-effects model is adopted to incorporate data hierarchies in the analysis of time-course data, where temporal and subject effects are both assumed to be random. The method starts with the clustering of genes by fitting the mixture model within the multilevel random-effects model framework using the expectation-maximization (EM) algorithm. The network of regulatory interactions is then determined by searching for regulatory control elements (activators and inhibitors) shared by the clusters of co-expressed genes, based on a time-lagged correlation coefficients measurement. The method is applied to two real time-course datasets from the budding yeast (Saccharomyces cerevisiae) genome. It is shown that the proposed method provides clusters of cell-cycle regulated genes that are supported by existing gene function annotations, and hence enables inference on regulatory interactions for the genetic network.

A genetic algorithm based method for deregulated electricity market dispatch

In a deregulated electricity market, optimizing dispatch capacity and transmission capacity are among the core concerns of market operators. Many market operators have capitalized on linear programming (LP) based methods to perform market dispatch operation in order to explore the computational efficiency of LP. In this paper, the search capability of genetic algorithms (GAs) is utilized to solve the market dispatch problem. The GA model is able to solve pool based capacity dispatch, while optimizing the interconnector transmission capacity. Case studies and corresponding analyses are performed to demonstrate the efficiency of the GA model.

Optimal dispatch of spinning reserve in a competitive electricity market using genetic algorithms

Ancillary service plays a key role in maintaining operation security of the power system in a competitive electricity market. The spinning reserve is one of the most important ancillary services that should be provided effectively. This paper presents the design of an integrated market for energy and spinning reserve service with particular emphasis on coordinated dispatch of bulk power and spinning reserve services. A new market dispatching mechanism has been developed to minimize the cost of service while maintaining system security. Genetic algorithms (GA) are used for finding the global optimal solutions for this dispatch problem. Case studies and corresponding analyses have been carried out to demonstrate and discuss the efficiency and usefulness of the proposed method.

Genetic algorithm based distance spectrum technique for performance union bound of space-time trellis coded OFDM

This paper derives the performance union bound of space-time trellis codes in orthogonal frequency division multiplexing system (STTC-OFDM) over quasi-static frequency selective fading channels based on the distance spectrum technique. The distance spectrum is the enumeration of the codeword difference measures and their multiplicities by exhausted searching through all the possible error event paths. Exhaustive search approach can be used for low memory order STTC with small frame size. However with moderate memory order STTC and moderate frame size the computational cost of exhaustive search increases exponentially, and may become impractical for high memory order STTCs. This requires advanced computational techniques such as Genetic Algorithms (GAS). In this paper, a GA with sharing function method is used to locate the multiple solutions of the distance spectrum for high memory order STTCs. Simulation evaluates the performance union bound and the complexity comparison of non-GA aided and GA aided distance spectrum techniques. It shows that the union bound give a close performance measure at high signal-to-noise ratio (SNR). It also shows that GA sharing function method based distance spectrum technique requires much less computational time as compared with exhaustive search approach but with satisfactory accuracy.

Optimal design of a regenerative dynamic dynamometer using genetic algorithms

This paper presents an approach for optimal design of a fully regenerative dynamic dynamometer using genetic algorithms. The proposed dynamometer system includes an energy storage mechanism to adaptively absorb the energy variations following the dynamometer transients. This allows the minimum power electronics requirement at the mains power supply grid to compensate for the losses. The overall dynamometer system is a dynamic complex system and design of the system is a multi-objective problem, which requires advanced optimisation techniques such as genetic algorithms. The case study of designing and simulation of the dynamometer system indicates that the genetic algorithm based approach is able to locate a best available solution in view of system performance and computational costs.

Zirconium alloying and grain refinement of magnesium alloys

Factors that influence alloying zirconium to magnesium with a Mg-33.3Zr master alloy and the subsequent grain refinement are discussed based on a large number of experiments conducted at the laboratory scale (up to 30 kg of melt). It is shown that the zirconium particles released from the Zirmax(R) master alloy must be brought into thorough contact with the melt by an appropriate stirring process in order to attain a good dissolution of zirconium. The influence of alloying temperature on the recovery of zirconium was found to be negligible in the range from 680 to 780 degreesC. An ideal zirconium alloying process should end up with both high soluble and high total zirconium in the melt in order to achieve the best grain refinement in the final alloy. The distribution of zirconium in the final alloy microstructure is inhomogeneous and almost all of the zirconium in solution is concentrated in zirconium-rich cores in the microstructure.

A new zirconium-rich master alloy for the grain refinement of magnesium alloys

A new zirconium-rich magnesium-zirconium master alloy (designated AM-cast) has been developed by the CRC for Cast Metals Manufacturing in collaboration with Australian Magnesium Corporation for use as a grain refiner for magnesium alloys that do not contain aluminium. This work describes the microstructural characteristics of this new grain refiner and its grain refining ability when added to different magnesium alloys under various conditions (alloying temperature from 680 °C to 750 °C; weight of melt from 1 kg to 150 kg and sample thickness from 7 mm to 62 mm). Owing to its highly alloyable microstructure, AM-cast can be readily introduced into molten magnesium at any temperature when assisted by a few minutes of stirring or puddling. Little sludge has been found at the bottom of the alloying vessel in these trials due to the fine zirconium particles contained in the master alloy. The recovery of zirconium is normally in the range from 40% to 60% with respect to 1% zirconium addition as the master alloy. It is shown that this new master alloy is an excellent grain refiner for aluminium-free magnesium alloys.