Veska Gancheva

Parallel performance evaluation of multithreaded local sequence alignment

Bioniformatics is area, demanding knowledge and skills for acquisition, storing, management, analysis, interpretation and dissemination of biological information. This scientific area requires powerful computing resources for exploring large sets of biological data. Biological sequence processing is a key for molecular biology. Smith-Waterman algorithm, based on dynamic programming has become the most fundamental in bioinformatics. The goal of this paper is to investigate the efficiency of parallel multithreaded implementation based on Smith-Waterman algorithm. Performance parameters have been estimated by the experiments using various data sets on two parallel computer platforms.

Massively parallel algorithm for multiple biological sequences alignment

In silico biological sequence processing is a key for molecular biology. This scientific area requires powerful computing resources for exploring large sets of biological data. Multiple sequence alignment is widely used method for biological sequence processing. The goal of this method is DNA and protein sequences alignment. This paper presents an innovative parallel algorithm MSA BG for multiple alignment of biological sequences that is highly scalable and locality aware. The designed MSABG algorithm is iterative and is based on the concept of Artificial Bee Colony metaheuristics and the concept of algorithmic and architectural spaces correlation. The metaphor of the ABC metaheuristics has been constructed and the functionalities of the agents have been defined. The conceptual parallel model of computation has been designed. The algorithmic framework of the designed parallel algorithm has been constructed.

Parallel performance evaluation and profiling of multiple sequence nucleotide alignment on the supercomputer BlueGene/P

Biological sequence processing is a key of information technology for molecular biology. This scientific area requires powerful computing resources for exploring large sets of biological data. Multiple sequence alignment is an important method in the DNA and protein analysis. ClustalW has become the most popular tool and implements a progressive method for multiple sequence alignment. The goal of this paper is to propose the performance evaluation of the efficiency of parallel multiple alignment on the supercomputer BlueGene/P for the case study of investigating viral nucleotide sequences and finding out consensus motifs and variable domains in the different segments of influenza virus A genome. Parallel performance evaluation and profiling of multiple sequence alignment have been performed on the basis of parallel program implementation based on ClustalW method and a local mirror database of all available isolates of the 8 segments of the influenza virus A. The molecular biology outcome of the experiments is that the consensus and the variable domains in Influenza virus A under investigation have been determined and output by utilizing the biological sequence alignment editor UGENE UniPro.

Functional testing of object-oriented software

The goal of functional testing is to prove that the software performs in conformance with its specification. Functional testing emphasizes the behavioral characteristics of objects by focusing on the interface that is the public, and defining, feature of a class. There are functional test cases for each individual method and functional test cases that treat the class as a unified component. An algorithm that guides the construction of functional test cases for a class is presented. That algorithm can easily be modified to produce test cases at a greater level of coverage. The input to the algorithm will be portions of the class specification that is produced to guide the design process as much as for the testing process.

An approach to convert scientific data description

Data obtained from the scientific experiments are heterogeneous and distributed. They are stored in different file formats so it is difficult to be integrated, analyzed, processed and visualised. Scientific experiments require effective and efficient data management. In this paper an approach to convert raw data obtained from scientific experiments into relational database using XML based description is proposed. The approach is tested and verified by database and web service implementations. The implemented web service provides authorized access to the experimental data. In order to provide a secure data transmission and data access the database is encrypted through an asymmetric 128-bit encrypting method implementation based on public and private key. A number of experiments using specific scientific data obtained from various experimental instruments are conducted to illustrate the usage of the proposed approach.

Exploring the efficiency of parallel bacteria foraging metaheuristics for job shop scheduling problem optimization

This paper suggest a model of parallel computation based on Bacteria Foraging metaheuristics for Job Shop Scheduling optimization problem. The parallel model is verified experimentally by Hybrid (MPI + OpenMP) programming implementations, run on heterogeneous computer cluster. The experimental study comprises parallel performance evaluation, parallel algorithm scalability analysis, profiling and evaluation of solution quality.

Grid resource management using lottery scheduling

The GRID infrastructure provides an aggregation of a wide variety of distributed resources for solving large-scale data intensive problems in various fields. The aim of this paper is to propose a method for grid resource management based on lottery scheduling. The lottery scheduling algorithm is implemented as an extension of GridSim simulator that integrates an additional strategy for planning and resource management.