A.Tonnazzini; "Blind Source Separation in Bioinformatics", 24.10,13:40
Faculty of Engineering and Natural Sciences
FENS SEMINARS
BLIND SOURCE SEPARATION IN BIOINFORMATICS
Anna Tonnazzini, ISTI-CNR, Pisa
Abstract: This talk will be subdivided into two parts. In the first part, the bioinformatics group which is active at ISTI since 2005 will be described in terms of resources, activities carried out, methodologies employed and collaborations.
In the second part, some examples of bioinformatics problems that can be formulated in the framework of Blind Source Separation will be considered.
Recently, Independent Component Analysis approaches has been successfully proposed for the analysis and the clustering of data from DNA microarray experiments. In this talk, the possible application of ICA and BSS strategies to the discovery of gene regulatory networks will be highlighted, with particular reference to the prediction of regulatory modules of microRNAs and their target mRNAs.
In automated DNA sequencing, the final algorithmic phase, referred to as basecalling, consists of the translation of four time signals in the form of peak sequences (electropherogram) to the corresponding sequence of bases. Unfortunately, these signals are subject to several degradations, among which peak superposition and peak merging are the most frequent. In this talk, it will be shown that basecalling can be formulated as a BSS problem, and a priori information and Bayesian estimation can be exploited to remove degradations and recover the signals in an ideal impulsive form.
The statistical analysis of microspectroscopy signals is fundamental to evaluate the pigment composition of the photosynthetic compartments of algae belonging to different taxonomic divisions, and then to perform algae classification and phylogenetic comparison. The algae spectrum is modeled as the linear mixture, with unknown coefficients, of the pigment spectra. In a fully Bayesian setting, both the algae mixture coefficients and the parameters of a Gaussian bands decomposition of the pigment spectra can be estimated on the basis of the alga spectrum alone.
Bio: Anna Tonazzini graduated /cum laude/ in Mathematics from the University of Pisa in 1981. In 1984 she joined the Istituto di Scienza e Tecnologie dell�Informazione �A. Faedo� <http://www.isti.cnr.it/> (ISTI) of the Italian National Council of Researches <http://www.cnr.it/> (CNR) in Pisa, where she holds a researcher position at the Signals and Images Laboratory. She cooperated in CNR special programs and CNR-University agreements for basic and applied research in Image Processing and Analysis, Computer Vision, Neural Networks and Learning, and is co-author of over 60 published papers. She was the ISTI responsible for the UE Project Isyreadet <http://www.isyreadet.net/home.htm> (2003-2004), dealing with the enhancement of ancient degraded texts and manuscripts. Currently, she is the ISTI coordinator for the CNR Project Computational Biology <http://www.cnr.it/commesse/Scheda_Commessa_Descrizione.html?co=1428>, and is an Associate Investigator for the Low Frequency Instrument Consortium in the framework of the ESA project Planck <http://www.rssd.esa.int/index.php?project=PLANCK>, Inverse Filtering and Independent Component Analysis Team. She is also a member of the Document Image Analysis Benchmarking <http://iit.demokritos.gr/cil/bench/> group.
October 24, 2007, 13:40, FENS G032