Events in 2011

Markov processes of imprecise probabilities
20 December 2011
Damjan kulj, Phd
University of Ljubljana

Markov processes are one of the most widespread approaches to modelling time evolution of random phenomena. They typically involve estimation of a large number of parameters, which often is only possible with a significant degree of uncertainty. Such estimations then allow for different sets of parameters. A unique parameter set is then only possible to obtain by adding additional assumptions without proper background in available information. This leads to overly precise results not reflecting the uncertainty in the input information.

The models of imprecise probabilities generalise the classical probabilistic models and allow the analysis in situations where probabilistic parameters are not known with certainty. The uncertainty in input parameters is then properly reflected in the results.

Photos: 1, 2, 3, 4, 5, 6, 7.

Workflows for quantitative mRNA expression analysis: strengths and criticalities
22 November 2011
Prof Raffaele A Calogero, Phd
University of Torino

The processing methods used to extract and interpret the information are an important aspect of dealing with the vast amounts of data generated from RNA-seq. Although the number of computational tools for RNA-seq analysis is constantly growing, their strengths and weaknesses as part of a complex analytical pipe-line have not yet been deeply investigated.

An overview on second generation sequencing technologies and on mRNA-seq analysis workflow will be given with a more detailed view on miRNA-seq data analysis.

Photos: 1, 2, 3, 4, 5, 6, 7, 8.

Parameter Inference in Bayesian Statistics
24 October 2011
Gregor Gorjanc, PhD
University of Ljubljana

Inference of model parameters in Bayesian statistics will be presented. This task is often computationally demanding, though feasible with the introduction of approximation (stochastic and deterministic) methods and advancement in computing power. In the first part, Bayesian inference will be illustrated using an example where explicit computation is possible. In the second part, linear mixed model parameters will be estimated using one stochastic (Markov Chain Monte Carlo, MCMC) and one deterministic (Integrated Nested Laplace Approximation, INLA) method.

REGISTER-BASED CENSUS 2011 - new achievement of Slovenian statistics
15 June 2011
Danilo Dolenc, MSc
Statistical office of the Republic of Slovenia

The 2011 Census of Population, Households and Dwellings (which is already being implemented) is the first census conducted by the Statistical Office of the Republic of Slovenia without fieldwork data collection, but by integrating data from many already existing administrative and statistical sources. Approximately 20 different sources will be used. For conducting the register-based census, the changes of the methodology, of the statistical processing and of IT support were necessary. Besides basic characteristics of the new census method (above all advantages and restrictions), the main part of the presentation will focus on methodological differences between statistical and administrative concepts (in case of household data) and on statistical data editing where the following had to be taken into consideration: step-by-step data integration, balance between full automation of statistical processes and quality of data, repeatability of the phases of the process and provision of traceability of all changes of data in the statistical process from input to the final database.

Photos: 1, 2, 3, 4, 5, 6.

Models for Survival in Matched Pairs
30 May 2011
Mette Gerster, PhD
University Southern Denmark

Co-twin control designs are useful in identification of causal effects and have been applied to a vast number of different research questions within the fields of epidemiology and psychology. The logic of the co-twin control design assumes that twins who have been brought up together are matched on early environmental factors. Moreover, twins have partly or fully identical genetic setups at birth, depending on their zygosity. This implies that a number of factors - that are usually diffcult to measure - are held constant in within-pair comparisons. Hence confounding from these factors is controlled for per design thus providing less biased estimates of effect. Aside from confounder control, the differences in genetic relatedness between monozygotic (MZ) (genetically identical) and dizygotic (DZ) twins (share on average half of segregating genes) can be used for the purpose of making inferences about the source of confounding, i.e. genetic or environmental confounding, respectively. Assuming an association in the unpaired analysis, genetic confounding would be indicated if the within-pair analysis showed a partial attenuation of the association in DZ twins and a full attenuation in the MZ twins. Similarly, a full attenuation in both DZ and MZ twins would be compatible with shared environmental confounding. Finally to support a causal effect of exposure the association would have to persist in both DZ and MZ twins.

An application of the co-twin control design will be presented using a Danish twin study of the effect of education on breast cancer incidence (Madsen et al, 2011). Furthermore, I will discuss which survival models are appropriate to use in this context. One possibility is to use a shared frailty model. However, inference in the frailty model requires independence between the frailty variable and the explanatory variables (education in the above example). We study how violations of this assumption affects inference for the regression coeffcients, and conclude that substantial bias may occur.

Instead, we propose making inference by means of a stratified Cox model (Holt and Prentice, 1974) and we demonstrate that this model gives unbiased estimates regardless of a possible dependence between the frailty variable and the explanatory variable.

References: Madsen, M, Andersen PK, Gerster, M, Andersen, A-MN (2011). Education and incidence of breast cancer: Does the association replicate within twin pairs?, Br. J. Cancer, 104, 520-523. Holt, JD, Prentice, RL (1974). Survival analysis in twin studies and matched pairs experiments. Biometrika, 61, 17-30.

Photos: 1, 2, 3, 4, 5.

A flexible parametric alternative to the Cox model
12 April 2011
Paul C Lambert, PhD
University of Leicester

The Cox model is the most popular method for the modelling of time-to-event data. The fact that it does not directly estimate the baseline hazard function is both an advantage and a disadvantage. I will describe a flexible parametric alternative to the Cox model. The flexibility comes from the use of restricted cubic splines to model the log cumulative hazard. I will discuss the general idea of the flexible parametric approach and in particular the modelling time-dependent effects. I will cover some useful extensions to the approach that illustrate the advantage of using parametric models for survival analysis. This will include using age as the time-scale, quantifying both absolute and relative differences and relative survival.

Photos: 1, 2, 3, 4, 5, 6.

Chasing the doped athletes - too much or too little statistics?
15 March 2011
Maja Pohar Perme, PhD
University of Ljubljana Slides: PDF

Photos: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10.

Computing relative survival and estimating net survival: what are the routinely used methods doing?
10 February 2011
Prof. Jacques Estève, PhD
University of Lyon

Historically, the relative survival was defined as the ratio of the observed survival of a group of patients to the survival of an "identical group" free of the disease. Since finding such a control group is a hopeless undertaking, this leads to several methods of computation of "relative survival", which were supposed to improve the comparability of the control group, but little work has been done up to now in relation to the question: "what are this computing proposals estimating?" and in particular: "are we able to estimate the net survival of the group of patients from these ad hoc calculation?".

Besides giving a short review of what the routinely used methods are doing, the purpose of this presentation will be to give further arguments for using a methodology that has the purpose of estimating a well identified population value, which could be compared between groups (from region to region or from country to country)

This later objective is especially relevant for programs of research, which look at variation of survival with time and geographical region such as EUROCARE or CONCORD. Although multivariate models including the appropriate covariates could provide a theoretical solution, they are difficult to use in practice. A simpler approach has been suggested by POHAR et al. and can be easily implemented. Scientists, working on comparative study on population survival, should consider using it. It has already been decided to use it for the French population survival data after a careful check of its performance.

Slides: PDF

Round table about popularisation of statistics
7 January 2011

Round table about popularisation of statistics will take place on Friday, 7 January 2011, at 12.00 at IBMI. Suggestions are welcome.

Photos: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15.


X

OPOZORILO : Pregledujete staro stran IBMI

Vsebine na strani so zastarele in se ne posodabljajo več. Stara stran zajema določene članke in vsebine, ki pa morajo biti še vedno dostopne.

Za nove, posodobljene vsebine se obrnite na http://ibmi.mf.uni-lj.si/