Judith Cohn, Ph.D.
Technical Staff Member, Bioscience Division
Los Alamos National Lab

Judith Cohn's CV (.pdf)

Title: Dynamics Perturbation Analysis of SCOP Domains

We have developed an algorithm that uses analysis of protein dynamics to predict functional sites. The algorithm performs an approximate version of Dynamics Perturbation Analysis (DPA), which can predict ligand-binding sites in protein structures (D. Ming, M.E. Wall. 2006. J Mol Biol 358:213). The present algorithm decorates the surface of the protein with test points, and uses approximate calculations of entropies to characterize the degree to which each point perturbs the protein's thermal vibrations. Residues near points that cause a large change in entropy are predicted to reside in functional sites. We used the algorithm to analyze more than 50,000 SCOP domains; and predictions were integrated with residue-conservation statistics obtained from the HSSP database. The analysis was performed using a flexible, distributed software architecture recently developed for this and other computationally intensive tasks.

Peter Haug, M.D.
Senior Informaticist for Intermountain Health Care and Professor
Department of Biomedical Informatics, University of Utah

Peter Haug's CV (.pdf)

Title: Data Mining, Clinical Modeling and the Future of Healthcare Computing

Abstract: The evolution of the modern electronic health record (EHR) has created an environment in which large amounts of medical data are collected reflecting the character of disease processes and the response of caregivers to these processes. This data has historically had an episodic character but now is rapidly becoming longitudinal in nature, allowing it to reflect the course of health over time. The presence of this data facilitates both the study of disease processes over time and the development of novel, experience-based approaches to support the delivery of care.

The data available is captured both as structured, readily-computable information and as textual reports, which require approaches base on natural language processing (NLP) paradigms for effective, computer use. In this presentation we will discuss our ability to extract this data and to use it to develop computable models of disease capable of informing care over time. The conjunction of data mining and artificial intelligence techniques holds promise to change the way we deliver, document, and evaluate the clinical care process.

Kirk E. Jordan, Ph.D.
Emerging Solutions Executive IBM Strategic Growth Business/Deep Computing



Eugene Myers, Ph.D.
Group Leader
Howard Hughes Medical Institute, Janelia Farms Research Campus (HHMI JFRC)

Andrey Ptitsyn, Ph.D.
Assistant Professor, Colorado State University

Andrey Ptitsyn's CV (.pdf)

Title: Life Works on AC Power

Multiple studies indicate that 10-15% of all genes in the hypothalamus and multiple peripheral tissues in mammals oscillate in a daily (circadian) rhythm. In our recently published studies we have applied three alternative algorithmic approaches to identify circadian oscillation in metabolically active peripheral tissues in mice and reported unexpectedly high number of oscillating genes. Our studies also detect no steady non-oscillation fraction of actively expressed genes. This leads to the conclusion that the accepted null- hypothesis in tests for gene expression periodicity is formulated on the unfounded assumption that all genes display a default steady-line expression. We propose a new approach that allows application of Digital Signal Processing (DSP) algorithms separately to each phase class of genes. Combined with Kolmogorov-Smirnov test this method identifies circadian baseline oscillation in almost 100% of all expressed genes. We conclude that such prominence of circadian oscillation in gene expression must be taken into account in all studies related to biological pathways. The importance of oscillation in signal transduction is demonstrated on the example of insulin signaling. This suggests that the loss of synchronization is likely to be among the causes or aggravating factors in metabolic disorders such as obesity and diabetes.

Heinrich Roder, D. Phil.
Chief Technology Officer, Biodesix, Inc., Colorado

Heinrich Roder's Biography (.pdf)

Title: Comparative Mass Spectrometry for Clinical Applications

Abstract: The direct use of comparative mass spectrometry as an unlabeled probe for the differentiation of disease states, for prognostic stratification of patients according to treatments, and for disease progression monitoring fasces special challenges intricately connected with the physics of mass spectrometry of biological molecules. In this presentation I will give an overview of statistical techniques specific to the comparative analysis of mass spectrometry data.
David S. Wishart, Ph.D.
Department of Computing Science and Department of Biological Science
University of Alberta

David S. Wishart's Biography (.pdf)

Title: Metabolomics: The Next Frontier for Bioinformatics?

Abstract: Metabolomics is a newly emerging field of “omics” research concerned with the high-throughput identification and quantification of the small molecule metabolites in the metabolome. Metabolomics is drawing considerable attention these days because it has the potential to substantially improve the speed and accuracy of many clinical tests and diagnoses. Metabolomics shares many of the same computational needs with other, much better established “omics” fields such as genomics, proteomics and transcriptomics. All four “omics” techniques require electronically accessible and searchable databases, all of them require software to handle or process data from their own high-throughput instruments, all of them require laboratory information management systems (LIMS) to manage their data, and all require software tools to predict properties, pathways, relationships or functions. Unfortunately, for metabolomics specialists, relatively few of these essential tools or resources exist. Fortunately, for bioinformaticians, this represents a wonderful opportunity to apply what they have learned from other “omics” endeavors to this newly emerging field. In this presentation I will highlight some of the efforts we, and others are making to bring modern bioinformatics practices to metabolomics – including the development of public metabolomic databases, the development of LIMS and the creation of software to interpret metabolomic data.