Tentative announcement: This course will be withdrawn if there is insufficient interest Three day course on Bioconductor (intermediate level) Instructor: Vincent Carey, Ph.D. March 5,6,7, 9am - 5pm each day Inn at Longwood, Boston Massachusetts 342 Longwood Ave, Boston MA, 02115 Tuition: $600 academic, $1200 commercial Registration form: http://www.biostat.harvard.edu/~carey/form08.pdf Questions: stvjc at channing.harvard.edu -- please do not post questions on this course to the list A block of sleeping rooms will be available at Inn at Longwood at approximately $189/night; contact 617 731 4700 after Feb 5 and mention "Bioconductor conference". This course provides a hands-on survey of Bioconductor tools for working with genome scale data. The material targets students with reasonable facility with R at the command line who wish to get acquainted with data analysis for various experimental paradigms. We will cover, among other things: - the MAQC experimental design and platforms - the oligo package and new facilities for dealing with affymetrix chips (expression and DNA) - illumina expression and SNP chip data - SQLite facilities for biologic metadata and platform annotation - the MLInterfaces package for supervised learning - the GGtools package for genetics of gene expression Students who successfully complete the course will be enabled - to transform raw outputs from affymetrix and illumina platforms into analyzable ExpressionSets or allied containers, - to apply various forms of statistical analysis to answer questions about differential expression and genotype effects in genome scale data - to use various annotation resources such as GO and KEGG to help interpret patterns in genome scale data using only transparent and fully open source software Requirements: * prerequisites: There will be very little background material provided on either R or the assays to be studied. We are focusing on working with digital artifacts of experiments (possibly retrieved from GEO, or from a core, to which we may apply some QA, or which we accept as valid numerical data). If you have no prior experience with R but are interested in the course, be sure to have read Dalgaard, "Introductory Statistics with R" (Springer) and/or the introductory material on www.r-project.org. * equipment: Every student must bring a reasonably modern laptop computer with a DVD drive or a USB port to allow installation of several GB of software and data. All software and data are supplied for windows machines so that all students have identical working environments. Mac or Linux laptops may be used, but students using these will be expected to have good mastery of their operating system so that the majority of students, who use windows, will not be distracted by idiosyncratic support requests. Format: Each major topic is addressed in a brief lecture. A handout is provided with specific exercises and hints/partial solutions. Students work independently or in teams to solve exercises; the module concludes with discussion of the solution. Tentative curriculum Day 1: * morning: four technologies in 'cooked' form - transcript profiling: affy, illumina - CHiP-chip (yeast) - SNP-chips + expression - aCGH + expression * mid-day: containers: structure, population, methods - arrays - gene sets - browser tracks * afternoon: workflow components I - capture - QA - preprocessing Day 2: * morning: workflow components II: annotation resources - SQLite representations of array and general metadata annotations - web services * mid-day: statistical analysis concepts - categorical methods - limma and other regularized methods - multiple comparisons * afternoon: exploratory tools: visualization, PCA, clustering Day 3: * morning: exercises: MAQC, spike ins, genetics of gene expression * mid-day: category and enrichment analyses; supervised learning (MLInterfaces) * afternoon: reports and audits; reproducible research - Sweave/odfWeave The information transmitted in this electronic communica...{{dropped:9}}