|
|
 |
BCB 691 - Faculty Seminar Fall 2003
Fridays 12:10 PM - E 164
Lagomarcino*
*
Note: class meets in 1104 Gilman on Oct. 24
| Aug 29 | Overview & 10' Presentations by BCB Faculty (see BCB Rotation Projects) | ||
| Drena Dobbs, instructor
ddobbs@iastate.edu
Chris Tuggle, An Sci Allen Miller, Pl Path Susan Carpenter, VMPM |
BCB 691 Overview
Rotation Projects Rotation Projects Rotation Projects |
||
| Sept 05 | 10' Presentations by BCB Faculty (see BCB Rotation Projects) | ||
| Chris
Minion, VMPM
Randy Shoemaker, USDA/Agron Drena Dobbs, GDCB |
Rotation Projects
Rotation Projects Rotation Projects |
||
| Modeling Metabolic and Gene Networks | |||
| Sept 12 | Jean Peccoud | GDCB/Pioneer Hi-Bred | Modeling Gene Networks |
| Sept 19 | Eve Wurtele | GCDB | Overview
&
METNET: Integration of Metabolomic, Proteomic and Transcriptomic Gene Expression Data |
| Sept 26 | Srinivas Aluru | CprE | A Faster Algorithm for Hierarchical Clustering of Microarray Data |
| Macromolecular Structure-Function: Analysis & Prediction | |||
| Oct 03 | Drena Dobbs/Haibo Cao | GDCB/Physics | Overview &
Coupling Computational Protein Structure Prediction with Experimental Structure-Function Analysis |
| Oct 10 | Honavar/Dobbs | ComS/GDCB | Discovery of Protein Sequence-Structure-Function Relationships |
| Oct 17 | Alex Travesset | Physics | Soft Condensed Matter: from Physics to Computational Biology and Back |
| Oct 24 | Xueyu Song
* Note room change this week only: 1104 Gilman |
Chem | Protein Crystallization and Protein-Protein Interactions |
| Oct 30 - Nov 1 | 2003
International Symposium on Modern Computing
Iowa State University Sessions: Computational Intelligence, Application Specific IT Infrastructures, High-Performance Computing, Grid Computing |
||
| Oct 31 | No Class - attend Symposium lectures | ||
| Modeling: Evolutionary and Physiological Processes; Macromolecular Structures | |||
| Nov 07 | Karin Dorman | Stat/GDCB | Overview &
Mathematical Modeling in Biology: Applicationsto Virus Evolution |
| Nov 14 | Adrian Sannier | IMSE | Interacting with Macromolecules in C6 Virtual Reality Cave" |
| Nov 21 | Michael Smiley | Math | Mathematical modeling of cellular dynamics in response to environmental cues |
| Nov 28 | No Class - Thanksgiving | ||
| Structural, Functional, and Evolutionary Genomics | |||
| Dec 05 | Dan Voytas | GDCB | "Transposable Elements in Genome Evolution" |
| Dec 12 | Volker Brendel | GDCB | Overview &
"Genome Informatics - Challenges and Opportunities" |
| Dec 19 | No Class - Finals Week | ||
"Modeling Gene Networks"
Interactions occurring in the living cell between populations of macromolecules are now sufficiently understood to model them with some level of realism. The structure and dynamics of these models is reviewed. We will see that these models can be used to analyze gene regulation networks, metabolic pathways, signal transduction cascades in a common theoretical framework. A number of computational challenges will be introduced. As a possible application, I will show how such models can be evolved in silico to analyze the properties of a breeding strategy.
"METNET: Integration of Metabolomic, Proteomic and Transcriptomic Gene Expression Data"
Large data sets detailing many aspects of the molecular composition of a biological sample are leading to changes in the ways we look at biology. Ideally, these data sets could be analyzed globally to determine how the cell is regulated. However, methods such as differential equations can only be used to calculate a limited number of reactions. More importantly, precise information on the kinetics of interactions within the biological system, and even what the actual interactions is mostly missing. MetNet is under-development publicly available software, designed to provide a framework for the formulation of testable hypotheses regarding the function of specific genes. Our ultimate aim is to use this tool to provide the basis for identification of metabolic and regulatory networks that control plant composition and development. MetNet integrates genome-wide mRNA, protein, and metabolite profiling data. The software is focused on Arabidopsis, but can be expanded for use with other species. Gene expression data can be explored with the aid of a metabolic and regulatory network map together with interactive graph display, visualization, statistical analysis, and modeling tools. The MetNet metabolic and regulatory network Arabidopsis map is in a database (MetNetDB) of information on regulatory and metabolic interactions. This information is derived from a combination of extant WEB databases (ARACYC, KEGG, BRENDA) and expert biologist input. Expression data can be viewed using the multivariate graphic capability of the statistical data visualization software (GeneGobi). Metabolic or regulatory flow in the network can be explored via FCModeler. FCModeler captures the input from MetNetDB, in graphical form, and enables the user to identify pathways between entities. Networks will be modeled and results interpreted using simple fuzzy cognitive maps. Challenges we are addressing are: the many unknown interactions; proteomics and metabolomics data contain information on only subsets of the protein and metabolites; eukaryotes are compartmentalized; and the kinetics of different interactions are vastly different and frequently unknown. MetNet-VR is a virtual reality research and teaching tool.
"A Faster Algorithm for Hierarchical Clustering of Microarray Data"
Clustering is often the first step towards analyzing gene expression data resulting from microarray experiments. Hierarchical clustering using Pearson correlation coefficient is a popularly used clustering algorithm that runs in O(n^3) time, where n is the number of genes. In this talk, I will show a geometric transformation of hierarchical microarray clustering and use it to derive an O(n log n) algorithm.
October 3
Drena Dobbs, Haibo Cao
GDCB, Physics
"Coupling Computational Protein Structure Prediction with Experimental Structure-Function Analysis"
Theoretical protein structure models are sometimes generated as a "last resort" when the structure of a protein cannot be solved using experimental approaches. Even when a protein is amenable to experimental structure determination, integrating computational analysis and modeling with experimental approaches can provide valuable insight into protein sequence-structure-function relationships. During the past 5 years, significant advances have been made in computational prediction of protein structures. For example, homology modeling can generate a viable structural model when the sequence of a target protein is „30% identical to that of a protein whose structure is available. Unfortunately, many important proteins have little or no detectable sequence similarity to any protein of known structure. In this talk, we will present two examples of work in progress. In the first, we will illustrate how computational analysis of structures in PDB can be used to help decipher the sequence correlates of a potential regulatory "proline switch" discovered using NMR spectroscopy by Andreotti et al. In the second, we will describe an algorithm developed by Ho et al. for "structural threading" which, in some cases, can detect structural similarities in proteins with little or no detectable sequence similarity (i.e., <10% identity). This threading approach can be used for predicting the structure of a protein from its amino acid sequence or for genome-wide screening to identify protein sequences compatible with a known structure (e.g., to identify "missing" components of signal transduction or regulatory pathways).
"Discovery of Protein Sequence-Structure-Function Relationships"
The effective use of increasing amounts of data from disparate information sources to explore specific scientific questions specific presents several challenges in bioinformatics. We will discuss our efforts to develop and systematically evaluate computational methods for discovering sequence and structural correlates of protein function. This work requires analysis of large datasets derived from multiple information sources (e.g., protein sequences, protein structures, protein-protein interaction data, gene expression data). We are developing a platform for rapid and flexible assembly of the relevant data sets from several sources and tools for automated and systematic exploration of protein structure and function using data-intensive approaches. INDUS (Intelligent Data Understanding Environment) (http://www.cs.iastate.edu/~honavar/indus.html), under development in our lab, includes key elements of principled approaches to addressing challenges of user, task, and context-specific information extraction and knowledge acquisition. INDUS implements an ontology-driven, query-centric, approach to data integration that allows users to impose their own semantics on disparate data sources to efficiently acquire knowledge from distributed data. Representative knowledge acquisition tasks that arise in exploration of sequence-structure-function relationships in proteins are being used to evaluate and refine the current prototype of INDUS. A specific example, the analysis and prediction of protein-protein interactions will be presented.
"Soft Condensed Matter: from Physics to Computational Biology and Back"
Soft condensed matter investigates the phases and properties of liquid crystals, polymers, lipids or proteins from the knowledge of their intermolecular forces. In recent years, the sophistication of experimental techniques provides precise quantitative tests for detailed theoretical models, and the gap between soft condensed matter and biophysics is rapidly closing down. In this talk, I will describe the physical properties of the cytoskeleton of red blood cells, networks of actin filalments, vesicles of certain lipids or systems of polymers and lipids, and present detailed models for predicting their structural and mechanical properties. I will also show some of the applications the research has for drug delivery or packing and encapsulation of bio-materials, and discuss implications and future promising developments for computational biology.
October 24
Xueyu Song
Chemistry
"Protein Crystallization and Protein-Protein Interactions"
I will give an overview of our research interests in protein crystallization and protein interactions and will also discuss protein-RNA interactions, solvation dynamics in proteins, and charge transport in DNA, etc.
For research description and publications, see: http://www.chem.iastate.edu/faculty/Xueyu_Song/
November 7
Karin Dorman
Statistics and GDCB
"Mathematical Modeling in Biology: Applications to Virus Evolution"
I will begin with an introduction to the field of mathematical biology (biomathematics) and the concept of mathematical models. I will then discuss the use of a few mathematical models to better understand some problems relevant to virus evolution, including (1) the detection of recombination in HBV, (2) the spread of drug-resistant HIV variants in an infected individual, and (3) the persistence of EIAV in infected horses. For references, see: http://www.biomath.org/dormanks.
November 14
Adrian Sannier
Industrial and Manufacturing Systems Engineering
"Interacting with Macromolecules in C6 Virtual Reality Cave
This talk will describe facilities in the Virtual Reality Application Center (VRAC) and how they can enhance our understanding of the structure and function of molecules. In particular we will look at the small subunit of a bacterial ribosome, covering the importance of the ribosome, what we hope to gain by studying it, and how virtual reality can help.
November 21
Michael Smiley
Mathematics
"Mathematical modeling of cellular dynamics in response to
environmental cues"
One way in which cells are known to interact with their environment is through biochemical signals, such as growth factors. Cells may proliferate and move in response to these environmental cues. We describe some recently proposed mathematical models for this type of cellular dynamics. These models are continuum models comprised of coupled systems of ordinary and partial differential equations. Basic tenants of the modeling will be discussed, and specific models related to angiogenesis and tumor growth will be presented.
Application to tumor angiogenesis
December 5
Daniel Voytas
Genetics, Development and Cell Biology
"Transposable elements in genome evolution"
Transposable elements constitute a significant fraction of most eukaryotic genomes. The success of transposable elements as genome colonists suggests that they have adopted strategies to populate host genomes without compromising host fitness. One such strategy is targeted integration, wherein mobile elements identify 'safe havens' in the genome for insertion. Analysis of the chromosomal distribution of mobile elements suggests that targeted integration is widespread. Bioinformatics approaches are being coupled with wet-lab experiments to understand targeting mechanisms and the diversity of targeting strategies employed by mobile elements.
Genome data ranging from gene structure annotation to associated information on gene expression and gene product function are now widely accessible and penetrate biological research on all scales. Genome informatics seeks to provide infrastructure and analysis tools to facilitate knowledge acquisition from these data. I will review work in this area in our group related to gene structure annotation in plant genomic DNA.
Relevant web sites:
| Fall 2002 speakers: Amy Andreotti, Adam Bogdanove, Dan Ashlock, Susan Carpenter, Dianne Cook, Xun Gu, Mei Hong, Howard Levine, John Mayfield, Les Miller, Roger Wise |
| Fall 2001 speakers: Dean Adams, Srinivas Aluru, Madan Bhattacharyya, Julie Dickerson, Karin Dorman, Xiaoqiu Huang, J, Dan Nettleton, Dan Voytas, Roger Wise, Eve Wurtele |
| Fall 2000 speakers: Srinivas Aluru, Dan Ashlock, Julie Dickerson, Oliver Eulenstein, Kai-Ming Ho, Rich Honzatko, Vasant Honavar, Allen Miller, Hal Stern, Eve Wurtele |