ABSTRACT: Pattern discovery is a fundamental problem in both computer science and molecular biology with important applications in locating regulatory sites. In its simplest form, the signal finding problem can be formulated as follows: given a sample of sequences and an unknown pattern/motif that appears at different unknown positions in each sequence, can we find the unknown pattern? Despite many studies, this problem is far from being solved: many motifs in DNA sequences are so complicated that we don't yet have good models or reliable algorithms for their recognition. Many existing pattern finding algorithms are implementations of different local search strategies. The shortcoming of these algorithms is that, for subtle signals, they often converge to local optima that represent random patterns rather than the real one. We describe recent combinatorial and machine learning approaches to
detecting very subtle signals that push the performance of motif finding algorithms to its theoretical limit.

Geoffrey Hinton
Professor of Computer Science and Psychology
University of Toronto
Learning Energy-Based Models of High-Dimensional Data

ABSTRACT: Many researchers have tried to model perception using belief networks based on irected acyclic graphs. I shall argue that this approach is probably misguided because of the difficulty of inferring posterior distributions in densely connected belief networks. An alternative approach is to use layers of hidden units whose activities are a deterministic function of the sensory inputs. The activities of the hidden units provide energy contributions, and the probability of each sensory data vector is defined via its energy by using a Boltzmann distribution. The problem of perceptual inference vanishes in deterministic networks, so perception is very fast and accurate once the network has been learned. The main difficulty of this approach is that maximum likelihood learning cannot be used. Maximum likelihood adjusts the parameters to maximize the probability of the observed data given the model, but this requires the derivatives of an intractable normalization term. I shall show how this difficulty can be overcome by using a different objective function for learning. The parameters are adjusted to minimize the extent to which the data distribution is distorted when it is moved towards the distribution that the model believes in.

Francisco Ayala
Professor of Biological Sciences
University of California , Irvine
Between Utopia and Hades: Genetic Engineering and Mankind's Future

ABSTRACT:  Molecular biology has made it possible to manipulate the genetic make-up of humans.  Four technologies are available for the purpose.  Two can be used primarily for therapeutic purposes: genetic counseling and genetic surgery.  Two others have been proposed for improving the genetic lot of mankind, or to satisfy the wishes of individuals: germinal selection and reproductive cloning.  I will evaluate these technologies from a scientific perspective and explore their ethical and social implications

Luis Villarreal
Professor of Molecular Biology and Biochemistry
University of California , Irvine
Smallpox Virus: Reemerging Threats from a Vanquished Enemy Drives Scientific Innovation

ABSTRACT:  Smallpox virus was the first, and so far, only, infectious disease to be eliminated from the planet by human vaccine intervention.  The word vaccine itself (from Italian for cow), was used to describe this first immunization.  This virus was responsible for more human deaths, by far, than any other calamity.  In this lecture, I will outline the history of our understanding of this virus and its effect on human populations and culture.  I will discuss how we came to understand much about the scientific and molecular basis of this most fascinating virus, and its remarkably intimate relationship with its human host, as well as how it was eliminated.  I will also outline why the world is again threatened by its possible use as a bio-weapon and how its potential might be magnified.  Because humans are no longer infected with, or immune to smallpox, and because we have no animal model that works well to study a hypothetically altered disease that is not currently upon us, traditional approaches are not sufficient.  This situation has created a need to develop new, rapid and powerful technologies to generate safe vaccines, as well as to develop anti-viral therapeutic compounds.  These technologies will require the merger of the previously separated disciplines of molecular biology, immunology, chemistry, nano-engineering and computer science.  I will present some examples of how the smallpox bio-terrorism threat is driving the creation of new technology that may also address any new, or emerging infectious disease.  Systemic approaches that solve traditionally difficult problems, such as identifying protein-protein interactions, phage display for protein solubility problems and high throughput micro fluidic screening of inhibitors of protein function will be discussed.

Laurie Zoloth
Professor of Ethics & Director, Program in Jewish Studies
San Francisco State University/Northwestern University
Difficult Freedoms:  The Turbulent Ethical Issues in Human Stem Cell and Genetic Research

ABSTRACT:  As molecular biology is poised to embark on new scientific terrain in the research on human stem cells, cell signaling and genomics, the American public has been thrust into the most divisive debate since Roe v. Wade and the use of human embryos in medical research.  The charge has been raised by prominent political leaders that, with genomics, cloning and stem cells, scientists are playing God, and are leading us toward a future both post-human and tyrannical, yet others have insisted that it is this very view that blocks the road to the significant therapeutic opportunities that only unfettered research can provide. What should be the role for faith, ethics and politics in basic research policy? What is an ethical response to the questions raised by some religious communities? Can science be fairly regulated? This presentation will explore the salient ethical arguments on both sides of the debate and suggest that a way forward, is by a rich and thoughtful public debate on our duties and responsibilities, rather than our rights.

David Haussler
University of California Presidential Chair in Computer Science
University of California , Santa Cruz
Computational Analysis of the Human and Other Mammalian Genomes

ABSTRACT:  Working drafts are now available for the human, mouse and rat genomes, and other mammalian genome sequences are on the way.  We discuss the current status of these genome sequences and some of the key bioinformatic analysis problems presented by this data, including the problems of assembling the sequence, finding the genes and other functional elements, and reconstructing the evolutionary history of the genomes. Recent comparisons between the human and mouse genomes have revealed that approximately 5% of the human genome appears to be more conserved with the orthologous regions in mouse than can be explained assuming neutral evolution.  Is this the portion of the genome under selection for specific functions?  How can we use computational comparative genomics to further pinpoint functional elements in the genome?  We look at some of these questions, and discuss how mammalian genomes can now be explored and compared interactively on the web, at:  www.genome.ucsc.edu.

Douglas Wallace
Professor, Department of Biological Chemistry;
Director, Center for Molecular and Mitochondrial Medicine and Genetics
University of California , Irvine
The Mitochondria:  An Informatics Paradigm for the Post-Genomic Era

No abstract available

Ian Lipkin

Professor of Epidemiology
Columbia University
The Pandora's Box Project:  An Odyssey in Pathogen Discovery

ABSTRACT:  The goal of this project is to establish rapid, sensitive methods for virus detection, and apply them in global networks for infectious disease surveillance and pathogen discovery.  Genome projects and high throughput methods for profiling gene statement using cDNAs and oligonucleotides have revolutionized biology by providing tools for simultaneous assessment of thousands of nucleic acid sequences. Although these methods have been extended to analysis of polymorphisms and gene statement patterns in single viral systems (for example, HIV, CMV), they have not been employed to define complex viral flora or to implicate specific viruses in disease. There are still no high throughput molecular systems with which to pursue viral surveillance. Furthermore, for many viral taxa, we lack the information required to define broad hybridization targets that will comprise the foundation of such systems. The Pandora's Box Project will address these needs by expanding viral sequence databases and designing microarray, flow cytometric and multiplex PCR assays for rapid molecular detection and quantitation of known, novel, and bioengineered viral pathogens.

C. Lee Giles
Professor, School of Information Science and Technology
The Pennsylvania State University
Future Directions in Information Preservation and Aggregation – Cite Seer?