|
|
 |
|
News archives News tips
Stat 416-516 Course Description: Microarrays have become a central tool in functional genomics research. This course will focus on many of the key statistical issues related to the use of microarray technology. Students will be introduced to two-color microarrays (including cDNA and oligo microarrays) as well as single-channel platforms (e.g., nylon membrane arrays, Affymetrix GeneChips). The role of blocking, randomization, and biological and technical replication in microarray experiments will be discussed with particular emphasis on design of single-channel microarray experiments, design of two-color microarray experiments, and design of microarray experiments with a factorial treatment structure. Analysis topics will include normalization methods for microarray data; mixed linear model analyses to identify differentially expressed genes; adjustments for multiple testing: methods for weak and strong control of the familywise type I error rate, false discovery rate methods; and other analysis strategies including clustering and hierarchical modeling of microarray data. STAT 416X will emphasize application and practical use of statistical methods for designing and analyzing microarray experiments. STAT 516X will include greater emphasis on current statistical research topics related to statistical design and analysis of microarray experiments. Students completing STAT 416X should be able to design and analyze their own microarray experiments and describe their work in a manner suitable for publication. Students completing STAT 516X should be able to provide expert advice on microarray experimental design and perform appropriate analyses in collaboration with biological researchers. All students should gain a sound understanding of the statistical principles important for good microarray experimental design and analysis. Questions: Contact Dan Nettleton at dnett@iastate.edu, Department of Statistics, www.public.iastate.edu/~dnett. Quantitative Biology Course Description This seminar is a survey of the biostatistical methods commonly used in ecological and evolutionary research. The goal of the course is to give students a general picture of what statistical methods are commonly used in evolutionary ecology, which methods are appropriate for which types of data, and to provide a general knowledge of how the methods work. It is a 'think-first, learn equations second' approach to biostatistics, to provide an intuitive feel for what methods should be used when. Topics range from basic univariate statistics (ANOVA, regression, correlation) to multivariate general linear models (MANOVA, MANCOVA, multivariate regression), to exploratory methods (principal components multi-dimensional scaling, UPGMA clustering, etc.) We will also discuss common problems biologists make in data analysis, and, time permitting, discuss several 'specialty' topics such as spatial statistics, resampling methods, and meta-analysis. While this seminar is NOT intended to be a substitute for multivariate courses in statistics, it IS intended to provide sufficient detail so that students may pick up a recent issue of Ecology, Evolution, etc. and be able to understand the statistical methods used in the articles. This way, students may more critically evaluate the literature, and hopefully, be able to use these methods in their own research as the need arises. Below is a brief description of the planned topics for discussion: 1. Dots in space, 1 equation, and a multi-way table: Present a general framework and purpose for biological statistics (i.e. visualizing and describing point patterns); describe multi-way tables for deciding how to select appropriate analysis for your hypotheses and data types 2. ANOVA, Regression and Correlation methods: Review ANOVA and regression: commonly-used ANOVA models, multi-way ANOVA, multiple regression, and ANCOVA 3. Inferential Multivariate Statistics: Describe MANOVA, multivariate regression and multivariate 'correlation' techniques (e.g., canonical correlation and two-block partial least squares) 4. Exploratory Multivariate Statistics: Present ordination (PCA, CVA, PCoord, MDSCALE) and clustering (UPGMA, WPGMA, K-means) methods 5. Common Statistical Errors: 1: 'Statistics is more than a p-value' (i.e. always plot your data) 2: What to do when you have a significant interaction term for ANCOVA 3: Why you shouldn't standardize for a covariate (like size) a priori 4: Maximizing your sampling effort (why smaller n from more regions is better than larger n from fewer regions) 6. The Mantel Test: Introduction to matrix comparison tests, and their relation to ANOVA and regression 7. Other Spatial Statistics: Correlograms, semi-variograms, and point patterns (random, dispersed, regular) 8. Meta-Analysis: Basic objective of meta-analysis, how to combine data from multiple studies, effect sizes, how all meta-analysis models are variations of weighted GLM 9. Resampling Methods: Introduction to randomization, jackknife, and bootstrapping methods 10. Log Linear Models: Logistic regression, chi-square, multi-way tables, and 'logistic' PCA (correspondence analysis) Enumerative combinatorics is an exciting and rapidly developing area of mathematics. It deals with counting elements in a set of discrete structures and investigating properties of the resulting numbers as well of the structures themselves. Especially since the 1960s, enumerative combinatorics has come to involve problems and/or tools from other areas of mathematics such as algebra, geometry and topology, as well as from other fields like computer science, statistics, chemistry and biology. This course will be a graduate-level introduction to enumerative combinatorics. Students will be expected to have taken at least one undergraduate course in combinatorics or graph theory or abstract algebra or linear algebra, preferably in most or all of the above. TOPICS we will discuss in this course will be chosen from among: generating functions; q-enumeration; combinatorial statistics; partitions; lattices and lattice paths; plane partitions; symmetric functions and Young tableaux; patterns in permutations; hypergeometric series and hypergeometric identities. We will also investigate connections between various approaches to enumerative problems, e.g. generating function proofs vs. bijective proofs. GRADING: There will be no tests or exams. Homework is the core of this course. The final grade will be based on biweekly homework assignments. If you have further questions about this course, feel free to email Alex at burstein@math.iastate.edu. His website is located at http://www.math.iastate.edu/burstein. Horticulture 537 Course Description Coralie Lashbrook from Horticulture is offering a plant stress biology course this Spring of particular value to interdepartmental students in stress or hormone labs. Lectures are prepared entirely from current high-impact literature and the course presents an excellent chance to learn about the multitude of stresses beyond your own. Mini-grant proposal writing experience, proposal peer review experience and student presentations augment the class. Please contact her with any questions you may have. The catalog description follows: Hort 537. Environmental Stress Physiology. (Same as Agron 537, Bot 537.) (3-0) Cr. 3. Alt S. Prereq: Bot 320 or equivalent; BBMB 404 or equivalent. Physiology and molecular biology of plant responses to environmental stress. Emphasis on the role of hormones and hormone interactions in governing stress responses. Lectures are prepared from journal papers that elucidate key mechanisms controlling responses to drought, flooding, salt, nutrient deficiencies, freezing, pathogens and herbivores. Plants studied include genetic model systems and crops of horticultural and agronomic value. Contact Info for Dr. Lashbrook: 251 Horticulture Hall, Tel: 515-294-3789; FAX: 515-294-0730; webpage: http://www.hort.iastate.edu/facultystaff/meetfaculty/lashbrook/index.php Computer Science 573 Course Description: This course might be of interest to BCB students who are interested in the application of Machine Learning approaches (that combine statistical and information theoretic methods with algorithmic techniques) for analysis of, and discovery of meaningful relationships from macromolecular sequence, structure, function, expression, and interaction data. Course description is included below. BCB students can ignore the official prerequisites (listed in the catalog description) and take the course if they have had elements of probability and statistics (e.g. Stat 330), Algorithms, Data Structures, and Programming (at least at the level of ComS 208/228 with some knowledge of Java), and some discrete mathematics (e.g., ComS 330). Com S 573. Machine Learning. (3-1) Algorithmic models of learning. Learning classifiers, functions, relations, grammars, probabilistic models, value functions, behaviors and programs from experience. Bayesian, maximum a posteriori, and minimum description length frameworks. Parameter estimation, sufficient statistics, decision trees, neural networks, support vector machines, Bayesian networks, bag of words classifiers, N-gram models; Markov and Hidden Markov models, probabilistic relational models, association rules, nearest neighbor classifiers, locally weighted regression, ensemble classifiers. Computational learning theory, mistake bound analysis, sample complexity analysis, VC dimension, Occam learning, accuracy and confidence boosting. Dimensionality reduction, feature selection and visualization. Clustering, mixture models, k-means clustering, hierarchical clustering, distributional clustering. Reinforcement learning; Learning from heterogeneous, distributed, data and knowledge. Selected applications in data mining, automated knowledge acquisition, pattern recognition, program synthesis, text and language processing, and bioinformatics.
GOT NEWS?
URL:
|