Sheng Zhong, a new faculty member in Bioengineering at UIUC, will be our first presenter. The abstract for his talk appears below:

Slide set [PowerPoint]

Current Topics in Computational Biology

Sheng Zhong, Dept of Bioengineering, UIUC

Computational Biology has evolved rapidly from a tool into a discipline. Huge amounts of genomic data, including genome sequence, microarray data and a lot of others, have been produced with the hope to understand fundamental questions related to human health.

Computational developments that utilize genomic data to interrogate human health related problems are lacking behind the speed of data generation. Better algorithms and computational tools on machine learning, scientific computing, intelligence system, data management and information retrieval are in great demand. How these algorithms and tools should be developed to interrogate the important biological and biomedical questions is the art that is open to discuss in this lecture.

Two biological questions and their related genomic data are going to be discussed. 1) How do stem cells work? 2) What causes cancer? In detail, can we computationally identify gene clusters based on gene expression profiles that best represent the functional relationship among genes? What is the gene regulatory network that initiates stem cell differentiation, or cancer formation? How do we decipher the relevant DNA sequence code so as to understand the genetic bases of cancer?

Background Reading:

1. BBC News on stem cells: http://news.bbc.co.uk/2/hi/science/nature/4155016.stm Also related stem cell news on the right panel of the browser.
2. Forbes's survey of venture capital on stem cells http://www.forbes.com/newsletter/2004/11/01/cz_sg_1101soapbox.htm
3. Science: Joining Forces for Brain Tumor Research http://www.sciencemag.org/cgi/content/full/309/5731/32b

Scientific Reading:
1. Microarray study on stem cells. http://www.sciencemag.org/cgi/content/abstract/298/5593/597
2. Brain tumor microarray study http://www.nature.com/nature/journal/v415/n6870/full/415436a_fs.html

Richard LeDuc, research programmer for the NIDA Center for Neuroproteomics at UIUC, will speak on the topic of "Top Down Proteomics."

Slide set [PowerPoint]

Abstract:

This talk will describe the rapidly emerging technique of top down proteomics and review the current state of bioinformatic tools and techniques for identifying and characterizing proteins based on MS/MS data from intact proteins.

Biography:

Mr. LeDuc is a research programmer in the Center for Top Down Proteomics and the NIDA Center for Neuroproteomics both here at the University of Illinois. He has previously worked in the bioinformatics unit of the W.M. Keck Center for Comparative and Functional Genomics and as a contractor to the U.S. Environmental Protection Agency. He is also a part-time graduate student working on a PhD in Biometrics.

Title: Mining coherent dense subgraphs across multiple biological networks (joint work with H. Hu, Y. Huang, J. Han
and X. Zhou)

Slide set [PDF]

Abstract: The rapid accumulation of biological network data translates into an urgent need for computational methods for graph pattern mining. One important problem is to identify recurrent patterns across multiple networks to discover biological modules.  We developed a novel algorithm, CODENSE, to efficiently mine frequent coherent dense subgraphs across large numbers of massive graphs. Applying CODENSE to 39 co-expression networks derived from microarray datasets, we discovered a large number of functionally homogeneous clusters and made functional predictions for 169 uncharacterized yeast genes.

Availability: http://zhoulab.usc.edu/CODENSE/

Biography: Mr. Xifeng Yan is a fifth-year Ph.D. student in the Department of Computer Science at the University of Illinois.  His research interests include data mining, structural/graph pattern mining, and their applications in database systems and bioinformatics.

Title: Automatically Generating Gene Summaries from Biomedical Literature (X. LING, J. JIANG, X. He, C.~X. ZHAI, Q.~Z. MEI, B. SCHATZ)

Slide set: PPT | PDF

Abstract: Finding information about a target gene from biomedical literature is a very common task that all biologists routinely perform. By using a literature search engine, in particular, PubMed, a biologist needs to spend considerable efforts reading the retrieved articles in order to locate the most relevant knowledge about the gene. In this work, we study how to automatically generate a summary from the literature for a target gene to make it easier for biologists to access the already-discovered knowledge. We present a two-stage summarization method, which involves first retrieving relevant articles and then extracting the most informative sentences from the retrieved articles to generate a structured gene summary. The generated summary explicitly covers multiple aspects of a gene, such as the sequence information, mutant phenotypes, and molecular interaction with other genes. We conducted experiments on Drosophila genes from FlyBase and a subset of Medline abstracts about Drosophila. The generated summaries are quite informative, indicating that our approaches are effective in automatically summarizing literature information about genes.

Biography: Xu Ling is a first-year Ph.D. student in the Department of Computer Science at the University of Illinois. Her research interests are mainly in bioinformatics, including gene regulation, biomedical literature mining, and genome sequence analysis.

Title: Algorithms for understanding gene regulation and its evolution

Slide set [PDF]

Abstract: With the completion (and anticipated completion) of genomes of a wide variety of organisms, there is now an abundance of sequence data that needs to be analyzed in order to extract useful biological knowledge. In many cases, several closely related species (such as different varieties of fruit flies) have been sequenced, with the explicit goal of cross-species comparison to facilitate analysis. Therefore, development of accurate and efficient algorithms to perform this multi-species analysis of genomic data has become imperative.

In this talk, we will explore some probabilistic approaches to the above-mentioned problem, and in particular, to the understanding of gene regulation. Probabilistic constructs such as Hidden Markov Models will be shown to have a natural application in this context. When combined with a realistic mathematical model of evolution, the HMM will present itself as a powerful framework for multi-species genome analysis, with provision to systematically deal with biological noise and missing data. Some new experimental results cataloguing evolution of cis-regulation will be presented.

Finally, the talk will chalk out potential research projects that address many pressing questions in bioinformatics today, within the probabilistic framework presented.
 
Biography: Saurabh Sinha received his Ph.D. in Computer Science from the University of Washington, Seattle, in 2002. He was a post-doctoral fellow at the Rockefeller University, New York, from 2002-2005, before joining UIUC this Fall. His interests include algorithms in bioinformatics, particularly on solving problems in gene regulation and evolution.

Title: Subspace Clustering for Microarray Data Analysis: Multiple Criteria and Significance Assessment (joint work with Lei Liu and ChengXiang Zhai)

Slide Set [PDF]

Abstract: The goal of microarray data analysis is to identify a subset of  genes whose expression levels rise and fall coherently under a subset of  conditions, that is, they exhibit fluctuation of a similar shape when  conditions change. Typically, clustering algorithms are first used to  group together genes with similar patterns of expression. After  clustering, people would usually inspect the clusters manually with  annotation of genes and assign functions to clusters. Two major challenges  are thus (1) to find gene clusters that are truly meaningful  biologically; and (2) to rank and prioritize the clusters in some  reasonable order for human inspection. We present a new clustering method that addresses both  challenges.

Bio: Hui Fang is a fifth-year Ph.D. Student in the Department of Computer  Science at the University of Illinois. Her research interests include  information retrieval and bioinformatics.

Title: Internet Health Monitors: Distributed Infrastructure for Measuring Population Health

Slide set: http://www.canis.uiuc.edu/~schatz/monitors/health.monitors.ppt

Streaming video: http://beespace.cs.uiuc.edu:8080/ramgen/schatz100705.rm (RealMedia Player required)

Abstract: Healthcare is the dominant item in modern economies, and rapidly growing as the population ages. Current health systems are oriented towards acute illness in hospitals -- there is no current infrastructure that can handle the coming scale of chronic illness in homes. New information technologies can support viable healthcare infrastructure. Health monitors can measure the status of people in their homes and automatically route them to the appropriate provider. Such an infrastructure can be supported on a national scale, by providing personalized interactions via computer that generate detailed databases of patient records. Local interactions track progress of chronic conditions, while global analyses of individual records discover similar cases to guide individual healthcare.
My laboratory is building health monitor prototypes. These systems perform adaptive question asking, using quality of life questionnaires to ask targeted questions to individual patients, customized to their current conditions. The patient answers build a structured vector for each individual, precisely describing their current health status. These vectors can then be statistically clustered, to determine cohorts of similar patients requiring similar treatments. This talk describes concepts of Internet Health Monitors and preliminary experiments, along with plans for large-scale clinical trials within my new department.

Bio: Bruce Schatz is Professor and interim head in the department of Medical Information Science at UIUC, currently an "empty" department following the retirement of Nobel laureate Paul Lauterbur. He is principal investigator with the UIUC BeeSpace Project (www.beespace.uiuc.edu) and is affiliated with the Institute for Genomic Biology and the College of Library and Information Science. He also led the CANIS project (www.canis.uiuc.edu) .

Title: 3D Model of Synchronous Calcium Signals in Ventricular Myocyte

Abstract: Synchronized calcium signaling (SCS) is characterized by high gradient near the t-tubule membrane and low gradient in the cytoplasm in the transverse direction, which enables ventricular myocyte to respond rapidly and forcefully to electrical and chemical stimuli. We developed a 3D continuum model to investigate the role of structural and functional cellular components in regulating SCS. The model currently includes: 3D geometry of a single t-tubule and its surrounding half-sarcomeres; spatially distributed L-type calcium channel (LCC), sodium calcium exchanger and calcium pump; calcium entry and extrusion across the sarcolemma membrane; and calcium diffusion and buffering by ATP, fluo-3 and troponin C. To solve the reaction diffusion system, the finite element method in space was used in combination with the finite difference method in time. Results suggest that both t-tubule structure and the spatially heterogeneous distribution of calcium handling proteins are important for SCS. The model predicts that two aspects of heterogeneous distribution are required: the concentration of calcium handling proteins in the t-tubule membrane to be ~6 times of that in the surface membrane; and the concentration of LCC, in the cytoplasmic end of the t-tubule, to be ~2.3 times of that in the surface membrane end. These results suggest that heterogeneous distribution of calcium handling proteins within the t-tubule may be required for SCS, when the sarcoplasmic reticulum (SR) is inhibited. It also provides a foundation for further studies on the effects of three-dimensional t-tubule geometry and ion channel distribution on calcium dynamics with and without SR.

Bio: Kathy Lu is a research assistant professor in the Bioengineering department at the University of Illinois. She is interested in working on simulation of calcium dynamics in cardiac myocytes using distributed reaction-diffusion system solver.

Title: "3D Medical Volume Reconstruction: Complexity and Medical Community Infrastructure Support"

Slide set [PDF]

Abstract: We address the problem of optimal registration decisions during 3D medical volume reconstruction and their impact on (a) anticipated accuracy of aligned images, (b) uncertainty of obtained results, (c) repeatability of alignment, and (d) computational requirements. The registration decisions include (1) image size used for registration, (2) transformation model, (3) invariant registration feature (intensity or morphology), (4) automation level, (5) evaluations of registration results (multiple metrics and methods for establishing ground truth), and (6) assessment of resources (geographically local or distributed computational resources and human expertise). Our goal is to provide data-driven mechanisms for evaluating the tradeoffs between accuracy of 3D volume reconstructions and registration variables. First, we present links between registration decisions and 3D reconstruction results in terms of accuracy, uncertainty, consistency and computational complexity characteristics. Second, we have built software tools that enable geographically distributed researchers to optimize their data-driven registration decisions by using web services and high performance computing (HPC) resources. The support developed for registration decisions about 3D volume reconstruction is available to the general community with the access to the NCSA HPC resources. Next, we illustrate the performance of our registration decision support system by considering 3D volume reconstruction of blood vessels in histological sections of uveal melanoma from serial fluorescent labeled paraffin sections labeled with antibodies to CD34 and laminin. The specimens are studied by fluorescence confocal laser scanning microscopy (CLSM) images. Finally, we discuss the complexity of building a web-enabled, web services based, data-driven, registration decision support system for 3D volume reconstruction.

Background: The content of the talk will be based on five journal papers (Journal of Microscopy, International Journal of Web Services Research, and EURASIP Journal on Applied Signal Processing) and four conference papers (SPIE on Medical Imaging and IEEE International Conference on Web Services). The talk will provide an overview of our three years of NIH funded work jointly with UIC.

Bio: Peter Bajcsy is a research scientist at NCSA, UIUC, and an adjunct assistant professor in the CS and ECE departments at the University of Illinois. He is interested in X-informatics problems, where X stands for biology, medical, health care, hydrology, sensor and instrumentation domain specific information processing issues (see http://www.ncsa.uiuc.edu/people/pbajcsy/).

Uriel Kitron, head of the Spatial Epidemiology Lab at UIUC, will speak Oct. 28 on the topic, "Spatial Epidemiology of Infectious Diseases."

Slide set [PDF]

Abstract: The spatial distribution of infectious Disease, particularly those  transmitted by insects and the zoonoses is highly heterogeneous, in large  part because of environmental determinants of vector and reservoir host  distributions. Tools such as geographic informations systems, satellite imagery and  spatial statistics have many applications for studies of vector-borne  diseases and zoonoses. The issue of scale and spatial and temporal  resolutions is of key importance for research and control of infectious  diseases. A brief overview of spatial epidemiology and tools for spatial analysis  will be followed by examples from the Americas, including West Nile virus  in Greater Chicago, Illinois, Lyme disease in the U.S. and Chagas disease  on the village level in Argentina.

Bio: Uriel Kitron is Professor of Epidemiology and co-Director of the  Center for Zoonoses Research at the College of Veterinary Medicine. He is  also affiliated with the Program in Ecology and Evolutionary Biology, the  Department of Community Health and the Center for Wildlife Ecology in the  Illinois Natural History Survey. His research includes epidemiological  studies of West Nile virus and Lyme disease in the US, malaria and  schistosomiasis in East Africa and Chagas Disease and dengue in South America.

Title: Dynamics of mammalian chromosome evolution inferred from multispecies comparative maps

Slide set [PowerPoint]

Reading: 7/22/05 Science article by Murphy, Larkin, et al. [link]

Abstract: Comparative genome analysis in mammals has advanced significantly as a result of progress in genome mapping and DNA sequencing. We have studied the genome organization in eight phylogenetically distinct species to address fundamental questions relating to mammalian chromosomal evolution and to reconstruct ancestral mammalian chromosomes. Pairwise comparisons of the human genome sequence (NCBI build 33) with the sequence-based maps of two representative species of Rodentia (mouse, rat) were performed using the GRIMM-Synteny approach. In addition, comprehensive RH-based maps from representative species of three orders of mammals, Cetartiodactyla (pig, cattle), Carnivora (cat, dog), and Perissodactyla (horse), were included in the analysis.

The combined sequence and RH map-based analysis was conducted with a new bioinformatics tool that allowed for the visualization of homologous synteny blocks and lineage-specific or reuse breakpoints between the genomes analyzed. Identification of a significant number of reuse breakpoints confirms and extends our previous findings. These results provide a clear exception to the Nadeau-Taylor random breakpoint model of chromosome evolution. Analysis of gene content in and around evolutionary breakpoint regions revealed a marked enrichment in gene density compared to the genome-wide average, thus suggesting that at least some evolutionary breakpoints are acted upon by natural selection. Additionally, 98% of the primate-specific breakpoints contain segmental duplications that often flank inverted chromosomal segments. Analysis of centromeres and telomeres showed that significant numbers of telomere positions are conserved among mammalian species whereas centromeres have evolved more dynamically.

Association between the reuse evolutionary breakpoints and positions of centromeres implies that breakpoint reuse preferentially occurs at the sites of ancestral centromeres or neocentromeres in independent lineages. In contrast to previous findings, our analysis of chromosomal rearrangements in distinct mammalian species suggests an increase in the chromosomal breakage rate over the evolutionary time. This observation is supported by the remarkable similarity of ancestral genome architecture in reconstructed ferungulate and boreoeutherian ancestors. Also, we found an association between positions of cancer-associated chromosomal abnormalities in humans and the positions of evolutionary breakpoints, thus leading to the conclusion that some cancers may be a consequence of the same evolutionary mechanism that is necessary for speciation.

Bio: Denis Larkin is a visiting assistant professor in the Department of Animal Sciences, at the University of Illinois at Urban-Champaign. He is working on mapping and comparative analysis of the cattle genome, identification of cattle SNPs, and comparison of mammalian genomes.

Title: Next steps in DNA sequence analysis

Slide set [PowerPoint]

Abstract: DNA sequence analysis is often considered to be a mostly solved problem. However, the four DNA bases contain much information that is not currently accessible to biologists. One of the most prominent outstanding problems in biological sequence data analysis is the decryption of the intergenic or regulatory DNA sequence. We have developed an application, degsuite, that uses conventional statistics together with optimized algorithms to discover regulatory DNA motifs by comparison of sequence from differentially regulated genes. We are using the motifs detected using this method to develop a machine learning solution for analysis and prediction of plant gene expression based on regulatory sequence. In parallel, we are exploring the use of nanotechnology-based sequencing for the analysis of large, unsequenced crop genomes. We have sequenced 7% of the soybean genome and 90% of the soybean cyst nematode genome using nanowell sequencing. The informatics challenges raised by the nanowell sequencing method will be discussed.

Bio: Matthew Hudson has an MA from Cambridge University and a PhD from Leicester University. He was a post-doctoral fellow at the University of California, Berkeley from 1998 - 2002, and Bioinformatics Scientist at the Torrey Mesa Research Institute (later Diversa corporation) from 2002-2004, where he managed genomics, database and Java develoment projects, before joining the Crop Sciences department at UIUC as part of the IGB initiative. His interests include developing and applying computational methods to understanding the large and complex genomes of plants, particularly crops.

Sandra Rodriguez-Zas leads the Laboratory of Statistical Genetics and Bioinformatics at the University of Illinois. She is also a co-Principal Investigator on the BeeSpace Project at UIUC.

Title: Applications of linear models to microarray studies

Abstract: The challenges of analyzing and interpreting microarray data are a consequence of the potentially multiple sources of variation and complex experimental designs.  I will present three examples of application of statistical tools to gene expression information. First, I will compare the outcomes from Restricted Maximum Likelihood and Bayesian approaches to characterize gene expression patterns in different tissues. Second, I will present clustering and semiparametric approaches to identify major gene expression trajectories across physiological stages. Third, I will use linear mixed effects models to compare results from two microarray platforms used to characterize healthy and diseased samples.

Bio: Sandra Rodriguez-Zas is an Associate Professor at the Department of Animal Sciences and is an adjunct professor at the Department of Statistics, University of Illinois at Urbana-Champaign. She obtained her Ph.D. in quantitative genetics at the University of Wisconsin-Madison where she worked on Bayesian analysis of longitudinal data. Her areas of expertise are statistical genomics and bioinformatics and her research interests include modeling of gene expression data, detection of quantitative trait loci and bioinformatics tools to characterize nucleotide and amino acid sequences.