Four 6 hours courses each will run in parallel on Sunday 5th December, from 9 to 12 am and from 2 to 5 pm
1. Competing risks and multi-state models: concepts, methods and software
Presenters: Ronald Geskus (Academic Medical Centre, Amsterdam, the Netherlands), Hein Putter (Leiden University Medical Centre, the Netherlands).
Abstract: Competing risks and its extension to multi-state models play an increasingly important role in the analysis of time to event data. The morning session is devoted to competing risks. For competing risks models, there is a lot of confusion with respect to the quantities that can be estimated and their interpretation. We explain the main concepts and their interrelationship: the independence assumption; cause-specific cumulative incidence; marginal hazard, cause-specific hazard and subdistribution hazard; competing risks as a multi-state model. We discuss regression models on cause-specific and subdistribution hazard, and on cause-specific cumulative incidence. We show how analyses can be performed with standard software. In the afternoon, the extension to multi state models is discussed. Concepts like transition intensities and transition probabilities are explained. Nonparametric estimation and regression models are considered, as well as methods to obtain predictions of future events, given the event history and possibly covariate values of a patient. With right censored and/or left truncated data, we show that it is possible to perform many types of analyses using standard software, using the same techniques as in the multi-state representation of the competing risks model. Participants are expected to be familiar with standard survival analyses with a single endpoint, i.e. Kaplan-Meier, Cox models, and the counting process style notation..
2. Genome-Wide Association Studies
Presenters: Andreas Ziegler (Universität zu Lubeck, Germany), Kristel Van Steen (Université de Liège, Belgium)
Abstract: This 1-day course aims at providing guidelines on optimal practice in designing, analyzing, and interpreting GWA studies, for those who would like to actively analyze GWA data. Knowledge and awareness about all steps of a GWA analysis, including data preparation, quality control or statistical analysis will be considered as this is crucial for a cost-efficient use of preciously collected data. Although GWA studies have been and are successful in identifying main effects, the findings cannot explain total genetic heritability. Up-scaled high-quality main-effects GWAs, -omics data integration and interaction analysis at genomic scale will probably elucidate part of the problem. Therefore, also state-of-the-art approaches for detecting gene-gene and gene-environment interactions will be discussed in detail. Covered GWA-specific topics include quality control of genotyping technologies, safe-guarding measures against false-positive results, meta-analysis, dimensionality reduction with minimal loss of information and speeding up computation time.
The target audience is advanced graduate students and postdoctoral researchers from biostatistics, bioinformatics and life sciences. Each attendee is expected to familiarize himself/herself with basic concepts in genetics. This can be easily achieved by taking the online course available with the textbook by Ziegler & König (2nd ed 2009).
3. Multiple Imputation and its Application
Presenters: Mike Kenward/James Carpenter (London School of Hygiene and Tropical Medicine, UK)
Abstract: Missing data is ubiquitous in biomedical research, and there is now a substantial literature on various approaches to analysis of partially observed datasets. Among these, Rubin’s proposal for multiple imputation (Rubin, 1987) is an attractive, because once the missing data have been imputed, inference proceeds semi-automatically by fitting the scientific model of interest to each imputed dataset and then combining the results using general rules.
While multiple imputation for multivariate normal data is now well established, there continue to be significant developments in both algorithms for multiple imputation and their application. For example, handling complex longitudinal surveys (Nevalainen, Kenward and Virtanen, 2009); handling hierarchical structure (Goldstien, Carpenter, Kenward and Levin, 2009), and improving the robustness of multiple imputation (Daniel and Kenward, 2009). Together with today’s computers, this means that multiple imputation can now be applied to a very broad range of biomedical datasets.
Judging by consultancy and other courses we have led, we anticipate strong demand for a course which will update participants on recent developments, equip them to use MI in a range of applications and outline future research directions.
4. Statistics for Biological Networks
Presenters: Ernst Wit (University of Groningen, The Netherlands), Veronica Vinciotti (Brunel University, UK), Vilda Purutçuoglu (Middle East Technical University, Turkey)
Abstract: Networks have become a new paradigm in social, technological and scientific discourse, e.g. social networking websites, the world wide web itself, genetic pathways, etc. This development has been accompanied by new theoretical insights in the mathematical nature of networks. In this course we shall focus on biological networks, which arise in the emerging field of systems biology. The idea is that the functional genome is a stable system, whose emergent properties cannot be described via more traditional gene-by-gene approaches.
The aim of this course is to describe, model and infer biological networks using real genomic data. The course will deal with a large variety of statistical techniques, such as sparse graphical models, state space models, Boolean networks, hidden Markov models and (stochastic) differential equations. We will consider gene transcription data (microarrays), but also proteomics and other types of modern high-throughput data, such as microRNAs, ChIP-chip or RNAi data. The course is aimed at statisticians and bioinformaticians with a good knowledge of standard statistical methods (MSc-level Statistics) and who are able to use the statistical package R. Students are expected to have an affinity for genetics, but no explicit biological knowledge will be required
Fees (Brazilian currency):
Full Member R$300
Full Non-Member R$350
SCC Member R$150
SCC Non-Member R$175
SCC Student R$75