Discussion Session Thursday May 15, 2008 1 Find out whos working - PowerPoint PPT Presentation

Discussion Session Thursday May 15, 2008 1

 Find out who’s working on bioinformatics at Miami  Find out who has tools and expertise at Miami that can be applied to bioinformatics research  Get biologists and non-biologists to talk to (and maybe even understand!) each other 2

 Two-hour session  Brief introduction of attendees  Biologists – state research problems that desire collaboration on  Non-biologists – give tools and expertise available for collaboration on bioinformatics/biology problems  Break up into informal discussion session, with facilitation by  Chun Liang and Quinn Li, botany  Valerie Cross and John Karro, computer science 3

 Expertise in biostatistics  Analysis of dose-related tumorigenic trends in the presence of treatment-related toxicity  Analysis of pharmacokinetic data, particularly, methods for testing the equivalence of the areas under concentration-time profile curves  Risk assessment  Inverse regression/calibration problems where the dose associated with a particular level of response is estimated and tested  Optimal design of experiments for simple compartmental models  Integration of model uncertainty in the generation of risk estimates 4

 Analysis and optimization of algorithms  Interested in developing efficient algorithms for finding similar sequences in genomic databases  Work with problems that have well-defined measure of similarity or difference between objects  Improve problem solutions that currently use too much memory or take too much time  Edit distance (number of operations to change one text/genomic string to another) 5

 Study how insect viral genes (esp. baculovirus and ascovirus) are regulated in insect cells  Baculovirus – would like bioinformatic prediction of which AATAAA used in certain processing  Ascovirus – would like bioinformatic search for particular stem loop structure, which could then be verified in lab 6

Ontology - a vocabulary that represents a set of concepts of a particular domain and the relationships between those concepts  Gene Ontology (GO) guarantees the consistency of the referenced biological concepts in different databases  Use to annotate genes in various databases  Annotations used to determine similarity between genes and gene products  Group has made various ontology software tools 7

QUOTA Multi-view FCA OntoSELF 1: lagging and leading strand elongation, 8 CDC2, DBP11, POL2

 Primary focus on computationally-based analysis of DNA and RNA sequences  Develop tools to help with analysis  Example: Working on identification of functional genomic regions through comparison of genomes from related species  Example: Developed tools for the estimation of neutral substitution rates on a local scale  Study structure of rates  Study effect on evolution of genomic structure 9

 Software engineering  Software risk management and assessment  Probabilistic risk assessment  Software design methodology  Experimental verification of software design methodology effectiveness  Visual programming languages 10

 DNA tiling microarrays  Massive data sets  Broad coverage of genome  Low signal/noise ratio  Want to extract statistically significant information to justify validation experiments in a wet lab  Seek collaboration from statisticians to develop appropriate statistics  Seek collaboration from computer scientists to effectively implement statistical and data processing algorithms 11

 Looking for collaboration on genomic sequence assembly and clustering  Work with expressed sequence tags (EST) from complementary DNA (cDNA)  How trace a given set of ESTs back to their original genes?  New technologies can now very quickly sequence enormous amounts of short pieces of cDNA  Want computational tools to do correct assembly and clustering 12

 Software development in C/C++/Fortran for numerical computation  Conversion of software for parallel computation  Application support for various physics and biophysics packages, e.g., ANSYS, Abaqus  Modeling and simulation of vascular systems  Geometric model generation  Flow solving  Data visualization 13

 Expertise is applied probability  Served on graduate committees in zoology  Helped graduate students with data analysis  Experience in  Analysis of variance  Markov chains  Hidden Markov Models 14

 Expertise in optimization and simulation of complex systems  Bioinformatics experience  Sequencing by hybridization  Clustering the avian-flu viruses (with Henry Wan)  Working with Chun Liang (Botany) and CSA colleagues to cluster Expressed Sequence Tags (ESTs) to identify genes for conifers  Would like to hear from other biologists with similar research, e.g., use of ESTs for gene identification and regulation 15

 Has taught classes in introductory statistics, regression analysis, and time series analysis  Extensive experience applying statistics in business, social science, and natural science  Time series analysis to study chemical concentrations of stream flows into Acton Lake  Applications of regression techniques 16

 Scientific programming, especially C++ and MATLAB  Parallel programs on cluster  Graphical user interfaces (GUI) for programs  Mathematical modeling  Digital image processing  Basic knowledge of variety of mathematical techniques 17

 Works in Michael Kennedy’s lab  Seek collaboration and support for  Nuclear Magnetic Resonance (NMR) data  Use of principal component analysis (PCA)  Use of partial least squares discriminant analysis (PLS-DA) 18

 Installation and configuration of bioinformatics applications on the cluster  IT infrastructure planning and support - servers, network, storage, etc.  Scripting (writing programs for cluster) and help with cluster batch system  Database creation and advice on use  General support of cluster users 19

 Principal expertise  Mathematical optimization (theory, algorithms, software)  Modeling of decision problems  Research interests  Reformulating mathematical problems for efficiency  Applications of optimization to data-fitting  Parallel processing in optimization  Optimal design of experiments  Areas of application (to date)  Crystallography, statistics, hydrology, econometrics, toxicology, engineering, ecology 20

 Knowledge of statistics useful in  Microarray studies (separating signal from noise, cluster analysis, missing data), image analysis  Clinical studies, forestry and wild life, public health  Specific statistical tools  Bayesian hierarchical modeling and Markov chain Monte Carlo (MCMC) algorithms  Spatial analysis (areal data and point-referenced data), including prediction and model checking 21