David Hart

Parallel Implementation and Performance of FastDNAml — A Program for Maximum Likelihood Phylogenetic Inference

This paper describes the parallel implementation of fastDNAml, a program for the maximum likelihood inference of phylogenetic trees from DNA sequence data. Mathematical means of inferring phylogenetic trees have been made possible by the wealth of DNA data now available. Maximum likelihood analysis of phylogenetic trees is extremely computationally intensive. Availability of computer resources is a key factor limiting use of such analyses. fastDNAml is implemented in serial, PVM, and MPI versions, and may be modified to use other message passing libraries in the future. We have developed a viewer for comparing phylogenies. We tested the scaling behavior of fastDNAml on an IBM RS/6000 SP up to 64 processors. The parallel version of fastDNAml is one of very few computational phylogenetics codes that scale well. fastDNAml is available for download as source code or compiled for Linux or AIX.

The emerging role of biogrids

Four biomedically oriented grid systems, ranging from SARS diagnosis to arthropod evolution, demonstrate the promise of grid computing in medical practice and biological science.

Least-squares fit of an ellipse to anisotropic polar data: Application to azimuthal resistivity surveys in karst regions

Abstract Polar plots of various types of anisotropic data are often approximated by ellipses and used by earth scientists as part of the interpretation process. FITELLIPSE, a code to calculate the orientation and values of the major and minor axes of a best-fit ellipse to anisotropic data, is written using Maple, a standard commercial software. A nonlinear statistical parameter is calculated to evaluate the goodness-of-fit. Application to azimuthal resistivity in karst of Indiana demonstrates the direction and degree of the anisotropy.

LINPACK performance on a geographically distributed Linux cluster

Summary form only given. As the first geographically distributed supercomputer on the top 500 list, the AVIDD facility of Indiana University ranked 50/sup th/ in June of 2003. It achieved 1.169 tera-flops running the LINPACK benchmark. Here, our work of improving LINPACK performance is reported, and the impact of math kernel, LINPACK problem size and network tuning is analyzed based on the performance model of LINPACK.

Advanced information technology support for life sciences research

The revolution in life sciences research brought about by the sequencing of the human genome creates new challenges for scientists and new opportunities for computing support organizations. This may involve significant shifts in computing support strategies, particularly as regards interacting with life sciences researchers who maintain a medical practice. This paper describes Indiana Universitys experience in a large-scale initiative in supporting life sciences research, as well as several strategies and suggestions relevant to colleges and universities of any size. Computing organizations and support professionals have many opportunities to facilitate and accelerate life sciences research.

High performance computing: delivering valuable and valued services at colleges and universities

Supercomputers were once regarded as being of very limited use - of interest to a very few national centers and used by a small fraction of researchers at any given university. As scientific research becomes more and more dependent upon management and analysis of massive amounts of data, advances in human knowledge will become increasingly dependent upon use of high performance computers and parallel programming techniques. Indiana University has undergone a transformation over the past four years, during which the capacity, use, and number of users of High Performance Computing (HPC) systems has dramatically increased. HPC systems are widely viewed as valuable to the scholarly community of Indiana University - even by those researchers who do not use parallel programming techniques. Economies of scale and vendor partnerships have enabled Indiana University to amass significant HPC systems. Carefully implemented strategies in delivery of consulting support have expanded the use of parallel programming techniques. Such techniques are of critical value to advancement of human knowledge in many disciplines, and it is now possible for any institution of higher education to provide some sort of parallel computing resource for education and research.

Parallel computing in biomedical research and the search for peta-scale biomedical applications

Publisher Summary The recent revolution in biomedical research opens new possibilities for applications of parallel computing. This chapter discusses three parallel applications useful in medicine and bioinformatics, PENELOPE, GeneIndex, and fastDNAml, in terms of their scaling characteristics and the way they are used by biomedical researchers. PENELOPE is a code that may be used to enhance radiation therapy for brain cancers. GeneIndex is a tool for interacting with genome sequence data when studying its underlying structure. fastDNAml is a tool for inferring evolutionary relationships from DNA sequence data. The parallel computing community will have its greatest impact on biomedical research only if significant effort is expended in reaching out to the biomedical research community. As the search for peta-scale applications in biomedical research continues, standard techniques in parallel computing may result in useful, and even potentially lifesaving, applications at more modest scales of parallelism. The ability of parallel computing techniques enables more interactive analyses of biomedical data.

GeneIndex: An Open Source Parallel Program for Enumerating and Locating Words in a Genome

GeneIndex is an open-source program that locates words of any length k specified by the user in a sequence. GeneIndex is useful for understanding the structure of entire genomes or very large sets of genetic sequences, particularly in finding highly repeated words and words that occur with low frequency. GeneIndex accepts DNA sequences in FASTA format input files, and performs computations and input/output in parallel. GeneIndex has been implemented on Linux, IBM AIX, and NEC SX-8, and is available with test data sets (the entire genomes of Drosophila melanogaster and Homo sapiens). The performance of the program scales well with processor count -- that is, as the number of processors increases, the processing time required decreases proportionally.

Distributed parallel computing using windows desktop systems

Like many large institutions, Indiana University has thousands of desktop computers devoted primarily to running office productivity applications on the Windows operating system, tasks which are necessary but that do not use the computers’ full capacity. This is a resource worth pursuing. However, the individual desktop systems do not offer enough processing power for a long enough period of time to complete large scientific computing applications. Some form of distributed, parallel programming is required, to make them worth the chase. They must be instantly available to their primary users, so they are available only intermittently. This has been a serious stumbling block: currently available communications libraries for distributed computing do not support such a dynamic communications world well. This paper introduces Simple Message Broker Library (SMBL), which provides the flexibility needed to take advantage of such ephemeral resources. Condor [1] offers an approach to managing jobs on scattered computing resources that is well suited to this situation; there is a Windows version of Condor, although it does not at the time of this writing provide support for parallel computing. There are other systems for managing jobs in a distributed environment, such as Globus [2]. SMBL addresses a different problem: performing extended computations using a continually changing collection of small computers. We could not find a sufficiently fault-tolerant and wellbehaved PVM [3] implementation for Windows. MPI [4] implementations expect the same machines at the end of a job as at the beginning. This is only reasonable, since these libraries are generally used on dedicated systems. DOGMA [5] supports the desired type of computing, but only for applications written in Java. SETI@Home [6] does not provide a general-purpose framework. SMBL enables parallel computing on sporadically-available desktop systems by introducing a server to keep track of the processing nodes and route messages between them. The SMBL server acts as a communications broker for processes associated with a particular parallel job running on many different processors. SMBL is designed to work with heterogeneous systems. It is not a part of Condor, but they work well together. In conjunction, they can be used to run parallel jobs on Windows computers in an opportunistic fashion, without interfering with the computers’ primary users. Available as open source, SMBL is scalable, flexible and robust enough for a highly constrained and highly dynamic distributed computing environment, using ephemeral resources for massive computations.