[Forum SIS] 2 seminars - July 19 - h 14.00 - seminar room, III floor, Dip. di Matematica, Politecnico di Milano

[Alessandra Guglielmi]

Il giorno 19 luglio, alle ore 14.00, nell'aula seminari al terzo piano, 
Dipartimento di Matematica

del Politecnico di Milano (via Bonardi, 9, Edificio La Nave, Campus 
Leonardo, Milano)

si terranno due seminari. I dettagli di seguito.

Cordiali saluti

Alessandra Guglielmi

-- 
Alessandra Guglielmi
Dipartimento di Matematica - Politecnico di Milano
tel ++39.02.23994641
e-mail:alessandra.guglielmi at polimi.it


############################################

Annalisa Cadonna, Institute for Statistics and Mathematics, WU Vienna

Title: Spectral Density Estimation for Multiple Time Series

The problem of estimating the spectral density function arises naturally 
in fields where information about frequency behavior is relevant and 
several related signals are recorded concurrently. For example, 
multichannel electroencephalography (EEG) records measurements of 
electrical potential fluctuations at multiple locations on the scalp of 
a subject. I will present a hierarchical Bayesian modeling approach to 
spectral density estimation for multiple time series, where the 
log-periodogram of each series is modeled as a mixture of Gaussian 
distributions with frequency-dependent weightsand mean functions. The 
implied model for each log-spectral density is a mixture of mean 
functions with frequency-dependent weights. In addition to accommodating 
flexible spectral density shapes, a practical important feature of the 
proposed formulation is that it allows for ready posterior simulation 
through a Gibbs sampling algorithm with closed form full conditional 
distributions for all model parameters. I will show results for 
multichannel electroencephalographic recordings, which provide the key 
motivating application for the proposed methodology, and present some 
extensions to non-stationary time series.



Andrea Cremaschi, Oslo Centre for Biostatistics and Epidemiology (OCBE), 
University of Oslo

Title: A Bayesian model for the study of drug-drug interactions

Recently, increasing effort is being devoted to the study of the 
simultaneous administration of two drugs to the same kind of cell 
culture. The outcome may be representative of synergistic or 
antagonistic behaviour in terms of /cell viability/. A primary goal of 
this paper is to establish a reference value representing the condition 
where the combined compounds do not interact, also called 
/zero-interaction/ level. A number of approaches have been proposed that 
define such quantity, and then compare it with the response obtained in 
combination, to establish the amount of interaction present in the 
experiment at the moment of sampling. However, these approaches rely on 
different modelling assumptions on the concentration-response curve and 
on the mechanism of action, and may provide conflicting outcomes in 
real-life situations. In order to overcome these issues, we interpret 
the viability experiment in a probabilistic framework, by modelling 
single-cell quantities of interest, and including the information 
relative to different exposure conditions. In particular, we propose a 
Bayesian regression framework for modelling the response surface of two 
drugs combined, and show its performance on a wide simulation study, as 
well as on a diffuse large B-cell lymphoma (DLBCL) high-throughput 
screening dataset, comprising more than 400 drugs combined with the 
standard-of-care drug Ibrutinib. Posterior estimates of the 
zero-interaction level and of the interaction term are obtained via 
adaptive MCMC algorithms.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20180716/0a01ce7c/attachment-0001.html>