Simon Tavaré
Title of talk: "Approximate Bayesian Computation and the evolution of tumours"

Abstract. Multi-scale agent-based models such as hybrid cellular automata and cellular Potts models are now being used to study mechanisms involved in cancer formation and progression, including cell proliferation, differentiation, migration, invasion and cell signaling. Due to their complexity, statistical inference for such models is a challenge. Here we show how approximate Bayesian computation can be exploited to provide a useful tool for inferring posterior distributions. We illustrate our approach in the context of a cellular Potts model for a human colon crypt, and show how molecular markers can be used to infer aspects of stem cell dynamics in the crypt.

Bianca L De Stavola
Title of talk: "Familial and socio-economic influences on foetal growth across three generations: the Uppsala Birth Cohort Multigenerational study, Sweden, 1915-2002"

Abstract. Size at birth is a key indicator of the health of the newborn and a predictor of subsequent morbidity and mortality over the life course. Foetal growth is known to be influenced by familial factors, with twin and family studies having attributed 30-40% of the variance in birthweight to genes (maternal and foetal). However parental size partly reflects social disadvantage from earlier generations. We will analyse unique data across three generations from the Uppsala Birth Cohort Multigenerational (UBCoS Multigen) Study to quantify the continuities in biological disadvantage that can be attributed to earlier social disadvantage. UBCoS Multigen includes social and demographic variables on males and females born in Uppsala during 1915-1929, and on their children and grandchildren, if born before 2003. Path analysis will be employed to partition the size at birth associations between grandparents and their grandchildren into genetic, generation-specific and intergenerational environment contributions. The role of missing data and the inclusion of the missing data mechanism in the model will be discussed. Comparisons with the results obtained from standard linear regression modelling will be drawn and some general conclusions suggested.
This is work in collaboration with Ilona Koupil (Karolinska Institute and Stockholm University) and David Leon (LSHTM).

Ørnuld Borgan
Title of talk: "Case-control studies: an overview from a methodological perspective"

Abstract. The use of various types of case-control studies has increased rapidly over the last decades. While there are less than 350 papers in the Web of Science from the period 1970-79 containing the phrase "case-control", this number has increased to about 38 000 for the first decade of this millennium. This demonstrates the key role that case-control studies play in modern epidemiology. The theory of case-controls studies has its roots in the pioneering work of Jerome Cornfield in the early 1950s, and over the last 60 years a number of important methodological developments have taken place. The talk will provide a review of these developments, focusing on methods where control sampling takes place within a well defined cohort. Some future methodological challenges will also be discussed.

Els Goetghebeur
Title of talk:  "The hopes and hazards with instrumental variables for causal
inference from observational data: of Pharmaco-epi and Mendelian randomization"

Abstract. Electronic health records are being gathered on a massive scale. Hard and software
manufacturers are merging and promise data mining tools that will help predict and intervene  to achieve better outcomes.  The hope is well alive, and enormous methodological progress is being made that can help deliver the causal effects.
Reliable work is not a matter of quick and dirty however. In observational studies, the challenge of adjusting appropriately for the necessary confounders - especially in the time-varying setting- can be formidable. This adds greatly to the appeal of the alternative instrumental variables approach which mimics randomization based inference. In practice, that seemingly simple approach must meet serious challenges of its own. First, justification of the instrumental variable properties is rarely obvious. Second, there is a plausible or at least workable causal model to propose and third unbiased inference to draw.  We develop such an approach when the goal is to model risks and discover interventions that might change them.   We study the effect of Cox-2 inhibitors on the risk of gastro intestinal bleeding in this light and find out how the lessons learned carry through when Mendelian randomization provides the `ideal' instrument.
The talk will refer to joint work with Manoochehr Babanezhad  and Stijn Vansteelandt.

Jonathan Marchini
Title of talk: "Genotype Imputation and the 1000 Genomes Project"

Abstract. Genotype imputation involves predicting untyped genotypes in genetic studies and has been widely adopted in the analysis of genome-wide association studies. I will give an overview of the methods used, describe the factors that influence their performance and illustrate their use on real data. I will also describe how genotype imputation methods are being used in the generation of the datasets from the 1000 Genomes Project and illustrate how data from this project can be used to analyze genome-wide association studies.

Wolfgan Huber
Title of talk: "Differential expression analysis for sequence count data"

Abstract. High throughput nucleotide sequencing provides quantitative readouts in assays for RNA expression (RNA-Seq), protein-DNA binding (ChIP-Seq), cell counting. Statistical inference of differential signal in these data needs to take into account their natural variability throughout the dynamic range. When the number of replicates is small, error modeling is needed to achieve statistical power. We propose an error model that uses the negative binomial distribution, with variance and mean linked by local regression, to model the null distribution of the count data. The method controls type I error and provides good detection power. A free open-source R software package, DESeq, is available from the Bioconductor project.