March 3, 2021, Wednesday


Jump to: navigation, search


BIMIB Monthly Seminar

The Monthly BIMIB Seminar is usually held every last Wednesday of the month at 14:00 in Building U14 of the Università di Milano-Bicocca. It is an occasion for researchers to talk about current activities in Bioinformatics and Systems Biology in an informal setting.

Upcoming BIMIB Seminars


Meetings, Workshops and Schools

15 Aprile 2019, h 9:00, Aula Martini U6-408, Edificio U6, Università degli Studi di Milano Bicocca
May 28-31, 2019, Villa del Grumello, Como, Italy.

Past Events

May 22-25, 2018, Villa del Grumello, Como, Italy.
  • Seminar: Population dynamics and optimal control: an application to the chronic myelod leukemia treatment.
Fabrizio Angaroni, Università degli Studi dell'Insubria,
Thursday, March 15th, 2018, 11:00, Room TBA, U14, DISCo Università degli Studi di Milano Bicocca, Milan, Italy.
Many tumors are hierarchically organized and driven by a sub-population of tumor initiating cells, or cancer stem cells. In the first part of this seminar we are going to show different a quantitative methods based on a mathematical models useful to describe a hierarchically organized tumor dynamics, in particular a differential equation method, a Montecarlo simulation and a finite difference probabilistic dynamics.
In the second part we are going to show some basic techniques of optimal control. The goal of such kind of study is to find an optimal feedback control strategy to minimize the "cost" of the therapy, where optimality means balancing some control costs against performance.
I am going to show how is possible to obtain an exact solution using the Pontryagins maximum principle, then an example of numerical resolution using an extension of Pontryagins maximum principle for discrete time systems and eventually an example of a random search algorithm: dressed chopped random basis algorithm (dCRAB).

  • Seminar: Tackling cancer complexity
Stefano Zapperi, Dipartimento di Fisica, Università degli Studi di Milano,
Caterina La Porta, Dipartimento di Scienze e Politiche Ambientali, Università degli Studi di Milano
Wednesday, February 28th, 2018, 11:00, Sala Seminari, First Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, Italy.
Recent advances on the plasticity of cancer stem cells are crucial to understand the metastatic process and guide therapeutic interventions. We will present experimental and modeling results describing these concepts in a quantitative way. On the other hand, we looked for a robust signature of obesity which would be able to highlight the connection with specific tumors related to this disease, such as breast cancer. We will also discuss these results, describing how we obtained a gene expression signature of obesity with a statistical significance of 5s.

May 23-26, 2017, Villa del Grumello, Como, Italy.

May 24-27, 2016, Villa del Grumello, Como, Italy.

Monday, May 23, 2016, 9:30-15:00, Sala Seminari, First Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, Italy.

  • Seminar: Regulation of self-renewal in cancer stem cells
Pier Giuseppe Pelicci, Istituto Europeo di Oncologia (IEO), and Department of Oncology and Haemato-oncology, University of Milan, Italy.
Wednesday, June 1st 2016, 11:00, Sala Seminari, First Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, Italy
Recent findings support the concept that cells with the properties of stem cells (SC) are integral to the development and perpetuation of several forms of human cancer, and that eradication of cancer stem cells (CSC) may be essential to achieve cancer cure. However, direct proof of these concepts is still lacking, mainly due the scarcity of appropriate model systems. We have recently defined a number of CSC-specific biological properties and underlying molecular mechanisms, using mouse models of i) leukaemia, obtained by transgenic expression of the PML-RAR, mutant NPM or AML1-ETO leukemia-associated oncogenes; and ii) mammary tumor, obtained by transgenic expression of the ErbB2 oncogene. We found that self-renewing divisions of CSCs are more frequent than normal counterparts, unlimited and symmetric, thus contributing to increasing numbers of SCs in tumoral tissues. SCs with targeted mutation of the tumor suppressor p53 possess the same self-renewal properties of cancer SCs, and their number increases progressively in the p53-null pre-malignant mammary gland. We showed that p53 signaling is attenuated in ErbB2-driven tumors, and that pharmacological re-activation of p53 induced restoration of asymmetric divisions in cancer SCs and tumor growth reduction, without affecting rates of apoptosis or proliferation on additional cancer cells. These data demonstrate that p53 regulates polarity of cell division in mammary SCs and suggest that loss-of-p53 in epithelial cancers favors symmetric divisions of CSCs, contributing to tumor growth. As a further mechanisms of extended self-renewal in cancer stem cells, we have demonstrated that up-regulation of the cell-cycle inhibitor p21 is indispensable for maintaining self-renewal of leukaemia SCs (LSCs). Expression of leukaemia-associated oncogenes in normal hematopoietic SCs (HSCs) induces DNA damage and activates a p21-dependent cellular response that, in turn, imposes cell-cycle restriction and triggers repair of the damaged DNA. This effect of p21 prevents the physiological exhaustion of HSC self-renewal, which occurs in time owing to accumulation of DNA damage, and confers an advantage to HSCs when they hyper-proliferate, as it occurs during stress or after full transformation (for example, in the LSCs), thus explaining the role of p21 in the maintenance of the self-renewal potential of LSCs. Finally, I will discuss unpublished data showing the contribution of immune-surveillance to the elimination of DNA-damaged SCs, and the underlying role of p21.
Joint work with Verga Falzacappa M.V., Insinga A., Tanaskovic O., Istituto Europeo di Oncologia (IEO), Milan, Italy.

  • Seminar: Modeling cumulative biological phenomena with Suppes-Bayes causal networks
Daniele Ramazzotti, Dipartimento di Informatica, Sistemistica e Comunicazione, Università degli Studi di Milano Bicocca, Milan, Italy.
Wednesday, February 24 2016, 11:00, Sala Seminari, First Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, Italy
Several statistical techniques have been recently developed for the inference of cancer progression models from the increasingly available NGS cross-sectional mutational profiles. A particular algorithm, CAPRI, was proven to be the most efficient with respect to sample size and level of noise in the data.
The algorithm combines structural constraints based on Suppes’ theory of probabilistic causation and maximum likelihood fit with regularization, and defines constrained Bayesian networks, named Suppes-Bayes Causal Networks (SBCNs), which account for the selective advantage relations among genomic events.
In general, SBCNs are effective in modeling any phenomenon driven by cumulative dynamical, as long as the modeled events are persistent. I will discuss on the effectiveness of the SBCN theoretical framework and its Bayesian interpretation.

  • Seminar: Docker new opportunities for bioinformatics, research institutes and enterprises
Raoul Bonnal, Istituto Nazionale Genetica Molecolare ‘Romeo ed Enrica Invernizzi’, Via F. Sforza 35, 20122 Milan, Italy.
Wednesday, January 27 2016, 11:00, Sala Seminari, First Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, Italy
Nowadays computing power is not a problem anymore, research labs can buy commodity hardware, build their own super computer or use cloud services. What is really challenging is how quickly those environments can be adapted to the different research needs and used efficiently. Some of those needs, such as reproducibility, versioning, maintainability, collaboration, openness can be tackled using Container technologies such as Docker and the adoption of the DevOps culture. To efficiently use the available computing power Docker can be combined with traditional “research” resource managers such as Torque, SLURM, SGE, LSF; use the cloud or more general resource managers such as Apache Mesos used by enterprises in heterogeneous environments. The seminar will show the benefits of converting bioinformatics pipelines in a series of containers; scratch the surface of Docker and its tools and the advantages of using Mesos as a resource manager.

  • Seminar: An introduction to RNAseq and its application to the analysis of extremely rare diseases
Daniele Merico, Informatics Core Facility Manager The Centre for Applied Genomics (TCAG), The Hospital for Sick Children, Toronto, Canada; Director of Molecular Diagnostics Deep Genomics Inc.
Monday, December 21st, 14:30, Room T023, Ground Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
The first part will be an overview of RNA-seq analysis: reads, splicing aware alignment, read counting and negative binomial models with variance shrinkage vs. isoform reconstruction or junctional counts. In the second part, I will present the application of RNA-seq analysis to confirming the molecular cause of an extremely rare congenital disorder, Roifman Syndrome, characterized by a defect of minor intron splicing resulting into widespread minor intron detection.

  • Seminar: Investigating the computational properties of biological neural networks
Thierry Nieus, Neuroscience Brain Technology, Istituto Italiano Tecnologia (IIT), Genoa, Italy
Friday, May 22nd, 2015, h 11:00, Sala Seminary, First Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
During my talk I will present some of my research activities going on at the Nets3 laboratory (IIT). In particular I will focus my talk on the computational properties of the cell culture dynamics recorded with the innovative multi electrode array recording system developed in the laboratory. Cell cultures are obtained from brain cells by retaining most of its biological properties and are therefore well suited to address biological (e.g. development, plasticity) as well as pharmaceutical (e.g. drug screening) questions. In addition, cell cultures can be grown to obtain variable topologies (e.g. 2D vs 3D, grid networks) and this can be of high interest to theoreticians. Moreover, the prominent activities displayed by cell cultures consist into propagating activities resembling those of many brain regions during their maturation phase. The latter further motivates the interest in developing data analysis tools and computational models to clarify the underlying mechanisms governing such emergent dynamics.

  • Seminar: Machine learning and in vivo imaging of the brain
Isabella Castiglioni, Institute of Molecular Bioimaging and Physiology of the National Research Council (IBFM-CNR), Milan, Italy
Wednesday, January 28th, 2015, h 11:00, Seminar Room, U14 DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
In many neurodegenerative and psychiatric disorders, early and accurate diagnosis is the main challenge in order to define effective and appropriate therapies.
Many scientific evidences prove that the diagnosis of such diseases, performed at different stages, can be supported by the integration of clinical examinations with in vivo measures of pathophysiological processes using molecular imaging techniques, such as Magnetic Resonance Imaging (MRI) or functional MRI (fMRI).
Until the last few years, methods used to analyse MRI-fMRI data were mainly based on mass univariate statistical approaches, e.g. Voxel-Based Morphometry [Ashburner 2000]. Recently, advances in statistical learning have led to a growing interest in multivariate pattern analysis (MVPA). This approach aims at the prediction of a single variable of interest (e.g. presence vs absence of a pathology) from the analysis and comparison of distributed patterns of brain anatomy or functionality. The main advantage of MVPA-based approaches is that, given their multivariate nature, they are able to detect cerebral patterns spatially distributed over a set of image voxels, and result in a relatively higher sensitivity than conventional univariate analysis [Mahmoudi 2012][Schrouff 2013][Pereira 2009].
Among MVPA approaches, Support Vector Machines (SVMs) [Vapnik 2000] [Shawe-Taylor 2004] showed promising results in the automatic classification of medical images, e.g. [Focke 2011], mainly for the classification of non-pathological vs. pathological subjects or for the classification of subjects belonging to different subtypes of disease.
The lecture is focused on the potential role of SVM in the classification of different neurological and psychiatric disorders (e.g. Parkinson, Alzheimer’s disease, eating disorders) by means of in vivo brain MRI studies. An SVM method tailored for the classification of brain MRI studies has been developed and validated. Current advantages and limitations of the proposed SVM approach will be discussed and presented.

Past Events (2014)

  • Seminar: Executable Heart: Computational Models and Methods for Detecting Emergent Behaviours in Cardiac Dynamics
Ezio Bartocci, Faculty of Informatics, T. U. Wien, Vienna, Austria
Tuesday, December 9th, h 11:00, Seminar Room, U14 DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Cardiac arrhythmia, such as atrial fibrillation (AF) and ventricular fibrillation (VF), is a disruption of the normal excitation process in cardiac tissue due to faulty processes at the ion-channel and cellular level, at the level of cell-to-cell communication, or at the full organ level. The clinical manifestation is a rhythm with altered frequency (tachycardia or bradycardia) or the appearance of multiple frequencies driven by spiral waves of electrical activity (polymorphic tachycardia), with subsequent deterioration to a chaotic signal known as fibrillation. These complex cardiac dynamics and instabilities can compromise the heart's ability to contract and to pump blood efficiently. In this talk I will present an overview of my research on computational models and methods for modeling and detecting such emergent behaviours in cardiac dynamics.

  • Seminar: Tackling complex inference tasks in large scale metabolic networks
Daniele De Martino, Center for life nanoscience-IIT, Rome
Thursday, December 11th 2014, h 14:00, Seminars Room, U14 DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Constraint-based modeling of metabolic networks is a promising framework for the inference of cellular metabolic capabilities from genomic data and rigorous physicochemical laws. In this talk I will give an overview of some computational methods, in particular Montecarlo markov chains, that we have recently developed for the implementation of thermodynamic constraints, the assessment of conserved chemical moieties and the uniform sampling of steady states.
Some applications to genome-scale metabolic network models will be discussed.
For information: Chiara Damiani chiara.damiani at

  • Seminar: Next Generation Sequencing: tra biologia ed informatica
Rocco Piazza, Dipartimento di Scienze della Salute, Università degli Studi di Milano Bicocca, ITALY
Friday, November 28th 2014, h 14:00, T023, Ground Floor, U14 DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Lo sviluppo delle tecniche di Next Generation Sequencing (NGS) ha causato un enorme incremento nelle nostre capacità di analizzare genomi complessi. Se solo alcuni anni fa il progetto genoma ha richiesto miliardi di dollari di finanziamenti ed anni di lavoro per il completamento del primo draft del genoma umano, oggi con un singolo strumento NGS è possibile sequenziare in una settimana 1000 miliardi di basi di DNA, equivalenti a circa dieci genomi umani con copertura 30x. Questo ha determinato una vera e propria rivoluzione nel modo di studiare patologie quali tumori o malattie ereditarie, consentendoci di spostare il baricentro dal singolo gene o da gruppi di geni all’intero genoma. Nel contempo queste tecnologie ci hanno messo di fronte a sfide fino a pochi anni fa inimmaginabili; se infatti in passato il collo di bottiglia di qualunque analisi genomica era rappresentato dalla produzione dei dati oggi lo scenario è esattamente opposto, con uno spostamento del collo di bottiglia dalla produzione dei dati genomici alla loro analisi. Ciò ha generato una serie di formidabili problemi di natura bioinformatica: dalla necessità di allineare in modo efficiente milioni di singole sequenze ad un genoma di riferimento a quella di sviluppare strumenti software in grado di analizzare in modo accurato ed efficiente Terabytes di dati di sequenziamento allineati, a quella di creare tool di data mining in grado di analizzare gruppi di esperimenti e di varianti geniche e di estrarre informazioni funzionali centrate su pathway.
Nel corso del presente seminario saranno prese in considerazione le interrelazioni tra le principali tecniche NGS e gli approcci bioinformatici applicati per la loro analisi.

  • Seminar: Formal analysis of biological models through HASL model checking
Paolo Ballarini, Laboratory MAS, Ecole Centrale Paris, Paris, FRANCE
Wednesday, May 28th 2014, h 11:00, Sala Seminary, First Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
When analysing a (stochastic) model of a biological system it is often desirable to assess specific (transient or steady-state) dynamical characteristics of an observed species, like the exhibition of peaks, the divergence/convergence as opposed to the existence of oscillations, or even the variability of the period/amplitude of oscillations, etc.
The Hybrid Automata Stochastic Logic (HASL) is a powerful formalism which provides the user with effective means for assessing sophisticated dynamical aspects of discrete-state stochastic models. Its essence consists in employing a hybrid automaton as a means for characterising the dynamical aspects that the trajectories of the observed model must fulfil (corresponding to the behaviour one wants to observe) together with specifying the quantities one wants to assess (wrt the observed behaviour). A (stochastic) simulation-based (statistical) procedure takes care of (automatically) sampling trajectories from the input model (which is given in terms of a Stochastic Petri Net) in a quantity sufficient to meet the desired accuracy of the output estimates. In this presentation I will demonstrate the potential of the HASL approach by looking at some biological applications. This includes 1) assessing the sustainment of translation vs transcription on a simple model of gene-expression 2) assessing the oscillatory characteristic of a model of the circadian clock 3) assessing the dynamics induced by a transient WNT signal in a model of the WNT pathway.

  • Seminar: Biology in mathematical models: a new strategy for brain investigation!
Egidio D'Angelo, Department of Brain and Behavioral Sciences, Unversit&agave; degli Studi di Pavia, Italy
Tuesday, Feb. 25th 2014, h 14:00, Sala Seminary, First Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Realistic modeling is an approach based on the careful reconstruction of neurons and synapses starting from biological details at the molecular and cellular level. This technique, combined with the connection topologies derived from histological measurements, allows a faithful reconstruction of neuronal networks. Finally, the advent of specific software platforms (PYTHON-NEURON) and of super-computers allows large-scale network simulation to be performed in reasonable time. This approach inverts the logics of original theoretical models, which anticipated an intuition on how the network should work. In realistic modeling, network properties “emerge” from the numerous biological properties embedded into the model. Since theoretical models have usually been unsatisfactory in predicting core central nervous system functions, and since these latter are in fact critically determined by biological details, realistic modeling provides the opportunity of bridging computational investigations to biology. The model is currently being used in the context of the HUMAN BRAIN PROJECT to investigate the cerebellar network and brain functions.

Past Events (2013)

  • Seminar: Tools for neuronal modeling
Paola Perin, Department of Brain and Behavioral Sciences, Unversit&agave; degli Studi di Pavia, Italy
Thursday, Dec 12th 2013, h 10:30, Room T-1403, Ground Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
I neuroni sono cellule specializzate per lo scambio di informazione. Le informazioni vengono codificate in segnali elettrici e trasmesse da un neurone all'altro a livello delle sinapsi. Per comprendere il funzionamento dei circuiti neuronali è quindi necessario conoscere i meccanismi che generano i segnali elettrici (canali ionici, regolazione della concentrazione intracellulare di calcio, neurotrasmissione chimica), le caratteristiche morfologiche delle strutture che li propagano e le connessioni tra neuroni. Diversi pacchetti software sono stati elaborati per simulare il comportamento elettrico dei neuroni; in questa lezione verrà esaminato soprattutto NEURON, uno dei principali standard. Alcuni modelli di singola cellula e di rete verranno mostrati per comprenderne il funzionamento e la logica di base.

  • Computational challenges in the analysis of high-throughput (epi)genomics sequencing data
Mattia Pelizzola, IIT, Milan, Italy
Wednesday, May 29th, 2013, h 15:00, Sala Seminari, I Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
In the last years, high-throughput sequencing methodologies have revolutionised the field of genomics, allowing profiling entire genomes with unprecedented coverage, quality and resolution. This seminar will cover the most important steps in the analysis of these data, and the application of these methodologies to the nascent field of epigenomics.

  • Identificare cosa? Le entità biologiche nell'era della Next Generation Sequencing
Maurizio Casiraghi, Dept. of Biotechnologies and Biosciences, Università degli Studi di Milano Bicocca, Italy
Wednesday, April 24th, 2013, h 11:00, Sala Seminari, I Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Il mondo della biologia sta affrontando una rivoluzione metodologica. Attraverso le tecniche collettivamente note come "Next Generation Sequencing" (NGS) è possibile analizzare finemente delle matrici complesse (acqua, terra, contenuto intestinale) nel quale numerosi organismi sono presenti contemporaneamente. L'approccio NGS prevede la combinazione di analisi molecolari e approcci bioinformatici, ma esistono due ordini di problemi: (1) le matrici di riferimento e comparazione sono largamente incomplete e molta biodiversità rimane di fatto non classificata. (2) la capacità di identificazione delle entità può essere influenzato dalla scelta del marcatore e da problemi intrinseci agli algoritmi utilizzati per la discriminazione nell'analisi bioinformatica.
Allo ZooPlantLab di Milano-Bicocca, Dip. Biotecnologie e Bioscienze, si affronta l'analisi di matrici complesse tramite NGS lavorando sia a un livello pratico, sia sulla parte concettuale del problema. Verranno esposte problematiche, lavori in corso e programmati.

  • Sox2 molecular targets in brain development and neural stem cells: from in vivo studies in mutant mice to genomic approaches
Silvia Nicolis, Dept. of Biotechnologies and Biosciences, Università degli Studi di Milano Bicocca, Italy
Wednesday, January 23rd, 2013, h 11:00, Sala Seminari, I Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Our project seeks to identify and functionally characterize molecular targets of the transcription factor Sox2, as mediators of its function in genetic brain disease. By Sox2 conditional ablation in mouse, we aim to identify genes regulated by Sox2, define Sox2 direct targets and their molecular mechanism of regulation by Sox2, and address their functional relevance for the pathological phenotype.
Sox2 heterozygous mutation in man causes a syndrome comprising a spectrum of nervous system defects. We previously reproduced essential aspects of this pathology by mouse genetic models of Sox2 deficiency. These showed that deficiency of one Sox2 target, the cytokine Shh, is critical for brain (hippocampal) pathology of these mice (also found in patients), as this was partially rescued by a drug mimicking Shh (Favaro R et al., Nat. Neurosci. 2009). More recently, we found that Sox2 ablation in early development (via a BF1-Cre transgene) causes major tissue loss in ventral telencephalon (striatum, GABAergic neurons), again depending on Shh; this reproduces aspects of holoprosencephaly, a disease that can be caused by SHH mutation. By a candidate-gene approach, we identified Nkx2.1, a known master regulator of ventral telencephalon development, and a direct regulator of Shh, as a novel Sox2 target gene. We also uncovered abnormalities in dorsal telencephalic (cerebral cortex/hippocampus) regions, that we are further characterizing.
We are also addressing Sox2 targets by genomic approaches. By RNAseq, we identified many Sox2-regulated genes by their differential expression in Sox2-ablated versus normal neural stem/progenitor cells (NSC). Many of these genes are bound, at their promoters, by Sox2. However, important DNA regulatory regions are often located at great distance from promoters of the regulated genes, and their location cannot be predicted based on proximity to the gene. We are addressing long-range interactions in the chromatin of normal, versus Sox2-ablated NSC by ChIA-PET (Handoko L et al., Nat. Genet. 2011). Sox2 ablation leads to a profound change in long-range interactions mediated by RNApolII, with both loss of many interactions, and generation of new ones. Many interactions involve genes differentially expressed by RNAseq, and many such genes are of known importance in the brain developmental processes impaired in our mutants. We are seeking to identify those long-range interactions that involve direct Sox2 binding (by available and new ChIPseq datasets), and will address their functional significance by transfection and transgenic studies.
The recent functional annotation of the human genome ( revealed that a large proportion of SNPs associated to genetic disease map within DNA regions of predicted regulatory significance (Maurano M et al, Science 2012). Identifying Sox2 targets will shed light on mechanisms of disease caused by Sox2, and may point to Sox2 involvement in diseases previously unrelated to Sox2.

Past Events (2012)

  • Mutation Profiles Reveal the Proliferation Dynamics of Colorectal Adenomas
Francesca Ciccarelli, PhD, Evolutionary Genomics of Cancer, European Institute of Oncology, Milan, ITALY
Wednesday, September 26th, 2012, h 11:00, Sala Seminari, I Floor, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY

  • NGS Milano Meeting
Tuesday July 17th, 2012, 14:30-18:30, Sala Seminari, I Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
The periodic meeting of the Milan area researchers and practitioners in the field of Next Generation Sequencing. The full program can be found here.

  • Seminar: A stochastic model of the emergence of autocatalytic sets
Alex Graudenzi, PhD, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Thursday, June 7th, 2012, 11:00, Sala Seminari, I Floor, U14, DISCo Università degli Studi di Milano Bicocca, Milan, ITALY
Autocatalytic sets (ACS) are rather common in biological systems and they might have played a major role in the transition from non-living to living systems. Several theoretical models have been proposed to address the experimentalists during the investigation of this issue and most of them describe a phase transition depending upon the level of heterogeneity of the primordial chemical soup. Nevertheless, it is well known that reproducing the emergence of autocatalytic sets in wet laboratories is a hard task. Thus, understanding the rationale at the basis of such a mismatch between theoretical predictions and experimental observations is a stimulating challenge.
In the talk a novel stochastic model of catalytic reaction network will be presented. The model is aimed at investigating the emergence of ACSs, sensibly considering the importance of noise, of small-number effects and the possible growth of the number of species and reactions in the system due to the overall dynamics.
Some important results regarding the generic properties of such a system will be illustrated, among which the quantitative analysis of the influence on the overall dynamics of: i) the composition of the incoming flux, ii) the composition of the initial set of molecules, iii) the average residence time of the molecules in the system. Preliminary studies concerning the introduction of energetic constraints and a parametric sensitivity analysis (PSA) of some key kinetic parameters will be also described.

  • Seminar: Fuzzy Tandem Repeats in DNA and Protein Sequences - algorithms and applications
Marco Pellegrini, Senior Researcher, IIT, C.N.R., Pisa, Italy
Tuesday, April 24th, 2012, h. 14:30, Sala Seminari, I piano, DISCo U14
Genomes in higher eucaryotic organisms contain a substantial amount of repeated sequences. Tandem Repeats (TRs) constitute a large class of repetitive sequences that are originated via phenomena such as replication slippage and are characterized by close spatial contiguity. They play an important role in several molecular regulatory mechanisms, in several diseases (e.g. in the group of trinucleotide repeat disorders), and in parental tests. While for tandem repeats with a low or medium level of divergence the current methods are rather effective, the problem of detecting TRs with higher divergence (fuzzy TRs) is still open. The detection of fuzzy TRs is propaedeutic to enriching our view of their role in regulatory mechanisms and diseases. Fuzzy TRs are also important as tools to shed light on the evolutionary history of the genome, where higher divergence correlates with more remote duplication events.
In this talk we will describe our approach to the detection of Fuzzy Tandem Repeats in DNA and in protein sequences, with application to symmetry detection in proteins. Moreover we formulate conjectures on a possible role for Fuzzy Tandem Repeats in the development of trinucleotide repeat disorders.
Joint Work with Alessio Vecchio and Elena Renda

  • Seminar: Approcci integrativi per identificare trascritti coinvolti nella progressione del carcinoma del colon-retto
Manuela Gariboldi, PhD, IFOM-IEO, Milan, Italy
Wednesday, January 25th, 2012, h. 11:00, Sala Seminari I piano, DISCo U14
Nel seminario verranno presentati dati sull'identificazione di trascritti (geni e miRNA) coinvolti nella progressione del carcinoma del colon-retto utilizzando un approccio di integrazione di dati di espressione e genomica. Questi dati sono stati ottenuti mediante analisi dell'espressione genica e di miRNA e di alterazioni del numero di copie di ciascun gene di casistiche di carcinomi del colon-retto e della corrispondente mucosa sana.

Past Events (2011)

  • Seminar: The Theorist's guide to the escape of tumors from immune control
Alberto d'Onofrio, PhD, Dept. of Experimental Oncology European Institute of Oncology, Milan, Italy
Wednesday, November 30th, 2011, h. 14:30, Sala Seminari I piano, DISCo U14
The competitive nonlinear interplay between a tumor and the host's immune system is not only spatio-temporally very complex but is also evolutionary.
A fundamental aspect of this issue is the ability of the tumor to slowly carry out processes that gradually allow it to become less harmed and less susceptible to recognition by the immune system effectors. Here we propose two simple epigenetic escape mechanism that adaptively depends on the interactions per time unit between cells of the two systems. From a biological point of view, our models are based on the concept that a tumor cell that has survived an encounter with a cytotoxic T-lymphocyte (CTL) has an information gain that makes it more fit than the naïve cells, which have never met a CTL. The consequence of this information increase is a decrease in both the probabilities of being killed and of being recognized by a CTL.
Numerical simulations of transitory phases complement the theoretical analysis.
Implications of the interplay between the above mechanisms and the delivery of immunotherapies are also illustrated.
[1] A. d’Onofrio and A. Ciancio, Simple biophysical model of tumor evasion from immune system control, Phys Rev E 84, 031910 (2011)
[2] M. Al-Taamemi, M.A.J. Chaplain, and A. d’Onofrio, Submitted
  • Seminar: A population-level model of Tumour-Immune System interplay: model construction and analysis
Giulio Caravagna, Dept. of Informatics, Systems and Communication, Università degli Studi di Milano Bicocca, Italy
Wednesday, October 26th, 2011, h. 14:30, Sala Seminari I piano, DISCo U14
The deterministic Kirschner-Panetta (KP) model for Tumour-Immune System interplay reproduces a number of features of this essential interaction, but it excludes the possibility of tumour suppression by the immune system.
In the first part of this talk we present a hybrid (i.e. stochastic/deterministic) version of that model able to reproduce the tumour suppression through stochastic extinctions when the KP model predicts oscillations.
In the second part we extend the hybrid model to account for anti-tumor therapies. We investigate the effect of varying therapy-related parameters (i.e. scheduling and administration) on the final outcome of the interplay between a tumor and the immune system. We perform this study via stochastic simulations of the hybrid model.
  • Seminar: Properties of membrane systems
Artiom Alhazov, Dept. of Informatics, Systems and Communication, Università degli Studi di Milano Bicocca, Italy
Wednesday, September 28th, 2011, h. 15:30, Sala Seminari I piano, DISCo U14
The talk will present a survey of some important properties of membrane systems that are referred to as "promising" or dynamic. It is sometimes quite desirable to achieve them, although they are mostly undecidable. We summarize how some of these properties affect the computational power or the descriptional complexity of membrane systems. A number of variants of the dynamic systems is also discussed.
  • Seminar: Set cover algorithms for very large datasets
Tony Wirth, Department of Computer Science and Software Engineering, University of Melbourne, Australia
Wednesday, July 27th, 2011, h. 12:00, Sala Seminari I piano, DISCo U14
The problem of Set Cover—to find the smallest subcollection of sets that covers some universe—is at the heart of many data and analysis tasks. It arises in a wide range of settings, including operations research, machine learning, planning, data quality and data mining. Although finding an optimal solution is NP-hard, the greedy algorithm is widely used, and typically finds solutions that are close to optimal. However, a direct implementation of the greedy approach, which picks the set with the largest number of uncovered items at each step, does not behave well when the input is very large and disk resident. The greedy algorithm must make many random accesses to disk, which are unpredictable and costly in comparison to linear scans. In order to scale Set Cover to large datasets, we provide a new algorithm which finds a solution that is provably close to that of greedy, but which is much more efficient to implement using modern disk technology. We also apply careful indexing techniques to implement the standard greedy algorithm. Our experiments show a ten-fold improvement in speed on moderately-sized datasets over a naive greedy implementation, and an even greater improvement on larger datasets.
  • Seminar: Gene-set association analysis of rare copy number variants in autism spectrum disorder and schizophrenia
Daniele Merico, Bader Lab & Emili Lab Donnelly CCBR, University of Toronto, Canada
Wednesday, July 13th, 2011, h. 9:30, Sala Seminari I piano, DISCo U14
Mapping rare genetic variants (such as genomic DNA copy number) is a new emerging strategy to study multigenic disorders that significantly impact reproduction fitness. Unlike common variants (typically SNPs), it is often hard to identify variants with significant association to disease, owing to the very low frequency even in the disease carrier population (e.g 1-3 variant carriers in a sample of 1000 disease carriers). For this reason, we addressed association testing at the gene-set rather than gene or variant level; significantly associated gene-sets will have a larger variant burden in disease carriers compared to controls. We will present results for rare copy number variants in autism spectrum disorder and schizophrenia, using gene-sets derived from functional annotations, pathway databases and protein interaction networks. Improved testing strategies taking advantage of gene interaction networks will also be discussed.
  • Seminar: Systems biology for cell signaling. Data-based mathematical modeling of VEGF signaling in quiescent and angiogenic endothelium as an example
Lucia Napione, Laboratory of Vascular Oncology, Institute for Cancer Research and Treatment – Candiolo (Torino), Dep. Oncological Sciences – University of Torino
Thursday, July 7th, 2011, h. 11:00 Sala Seminari Demografia, II piano U7, Dipartimento di Statistica.
There is a general agreement that a systems biology approach is needed for a better understanding of causal and functional relationships that generate the dynamics of cell signaling. These observations have been the basis for efforts to     establish tight interdisciplinary collaboration between biologists, mathematicians, physicists and computer scientists. Both quantitative wet-lab data and mathematical modeling are key features in systems biology. We have recently developed a data-based mathematical model of Vascular Endothelial Growth Factor-A (VEGF) signaling that will be used as an example to illustrate a systems biology approach to signal transduction and the insight that could be gained through modeling. VEGF is the major determinant for the activation of the angiogenic program leading to the formation of new blood vessels, a crucial event in tumor growth. VEGF specific binding to VEGF receptor-2 (VEGFR-2) triggers different  signaling pathways leading to endothelial proliferation, permeability and survival.
We carried out a quantitative in vitro analysis of these pathways by studying cultures of long-confluent and sparse endothelial cells that, at different VEGF doses, mimic the in vivo conditions of quiescent and angiogenic endothelium. We performed accurate quantitation of the activation levels of key proteins involved in VEGF-induced signal transduction and integrated the results by means of a mathematical model that satisfactory reproduces observed data and shows a predictive potential.
The principal finding of this study is the formal demonstration of the influence of endothelial cell density on the activation of the early triggered signaling events along VEGF/VEGFR-2 axis. Our results provide a basic framework suitable for further extensions that will shed light on the complexity of the VEGF signaling and prompt for future investigation aimed at identifying molecular determinants responsible for the fine regulation of VEGF signaling.
  • Seminar: From cellular to molecular Growth & Cycle models
Marco Vanoni, Dept. of Biotechnologies and Biosciences, Università degli Studi di Milano Bicocca, Italy
Wednesday, May 25th, 2011, h. 11:00, Sala Seminari I piano, DISCo U14
The seminar will summarize the major connections between cell growth and cell cycle in the model eukaryote Saccharomyces cerevisiae. In S. cerevisiae regulation of cell cycle progression is achieved predominantly during a narrow interval in the late G1 phase known as START. At START a yeast cell integrates environmental and internal signals (such as nutrient availability, presence of pheromone, attainment of a critical size, status of the metabolic machinery) and decides whether to enter a new cell cycle or to undertake an alternative developmental program. Several signalling pathways, that act to connect the nutritional status to cellular actions, will be briefly briefly outlined.
A Growth & Cycle Interaction Network has been manually curated. More than one fifth of the edges within the Growth & Cycle Network connect Growth and Cycle proteins, indicating a strong interconnection between growth and cycle. The backbone of the Growth & Cycle network is composed of middle-degree nodes suggesting that it shares some properties with HOT networks. A cellular model has been developed that integrates cell growth (described by accumulation of of ribosomes and proteins) and cell cycle of a yeast cell. Each steady state condition of growth is characterized by a given level of Ribosome/Protein. A negative feedback (able to detect inactive ribosomes) controls ribosome synthesis. A cell sizer threshold involving Cln3 and Far1 controls entrance into cycle. The G1 phase is divided into two periods: timers T1 (starting after Threshold overcoming) is different in P and D cells, timer T2 is identical in P and D cells. The S phase starts at the end of timer T2, coincident with initiation of budding timer TB.
At the end of the cycle, a resetting function endows both P and D cells fn appropriate cell mass (D cell take all biomass synthesized after budding initiation) and the same set of SPI proteins. The interplay of these SPI proteins and cell mass sets cycle dynamics of both P and D cells. Increasing genealogical age generates a mechanical stress that reduces the rate of net synthesis of protein accounting for increase in size of P cells.
The model describes growth of single cells in both fast and slow conditions. Experimental findings are used to set the parameter values at a fast growth rate (MDT = 100 min) and at a slow growth rate (MDT = 180 min). The model has been extended at the population level. It allows to describe yeast populations (in terms of budding, macromolecular composition, increase in cell number) both during exponential growth and during nutritional shifts.
The structure of the model allows to plug-in modules describing with increasing granularity molecular events of the cell cycle (just to name a few, the G1/S transition, DNA replication or mitotic exit) or signal transduction pathway by appropriately connecting input and output of the scaffold and plug-in modules. Experimental and computational work along this line will be presented.
  • Seminar: Algorithmic models for DNA processes
Giuditta Franco, Department of Computer Science, University of Verona, Italy
Wednesday, May 25th, 2011, h. 14:30, Sala Seminari III piano, DISCo U14
A recent trend in DNA computing research is concerned with innovative DNA procedures of interest in molecular biology. In fact, new bio-techniques may be designed as efficient combinatorial algorithms on biological strings. As an example of such an approach, a few recent DNA recombination algorithms and experiments will be presented in this talk, together with a variant of the Polymerase Chain Reaction (called XPCR) technique, which is a laboratory implementation tool of null context splicing rules.
If time permits, DNA (and bacteria) experiments in progress to prove the genetic drift emergence will be shown as well.
The talk will be as self-contained as possible in terms of both biological and formal language theoretical concepts; curiosity will be then the only requirement for attendance.
  • Seminar: An overview of modelling of intestinal crypt dynamics with references to colon-rectal cancer (CRC)
Marco Antoniotti and Giovanni De Matteis, BIMIB, Dept. of Informatics, Systems and Communications, Università degli Studi di Milano Bicocca, Italy
Wednesday, March 30, 2011, h. 14:30, Sala Seminari DISCo U14, first floor
Colon rectal cancers (CRC) are the result of sequences of mutations which lead the intestinal tissue to develop in a carcinoma following a "progression" of observable phenotypes. The actual modeling and simulation of the key biological structures involved in this process is of interest to biologists and physicians, and, at the same time, it poses significant challenges from the mathematics and computer science viewpoints. In this seminar we give an overview of some mathematical models for intestinal crypt dynamics and related problems and open questions. In particular, major attention is devoted to the study of the so-called lattice (or grid) and off-lattice (off-grid) simulation models. The current work is the groundwork for future research on semi-automated hypotheses formation and testing about the behavior of various actors taking part in the adenoma-carcinoma progression, from regulatory pathways to cell-cell signalling phenomena.
  • Seminar: Motif Discovery from Chips to ChIPs
Dr. Giulio Pavesi, Bioinformatics, Evolution and Comparative Genomics Lab, Dept. of Biomolecular Science and Biotechnology, Università degli Studi di Milano, Italy
Wednesday, February 2, 2011, h. 14:30, Sala Seminari DISCo U14, first floor
Motif discovery, aimed at the in-silico prediction of candidate transcription factor binding sites in nucleotide sequences has been, and still is, one of the most challenging and open research topics in bioinformatics. This talk aims to provide a survey of different approaches and methods to the problem, from the days when typical analyses were performed on sets of promoters from co-expressed genes obtained from microarray expression data (the Chips), to the latest advances in the field, that permit a more reliable identification of target sequences through genome-wide experiments like Chromatin Immunoprecipitation (the ChIPs).

Past Events (2010)

  • Seminar: A dynamical model of genetic networks describes cell differentiation
Prof. Roberto Serra, Deparment of Social, Cognitive and Quantitative Sciences, Università degli Studi di Modena e Reggio Emilia, Italy
Friday, November 26th, 2010 h. 14:30, Sala Seminari DISCo U14, first floor
Cell differentiation is a complex phenomenon whereby a stem cell becomes progressively more specialized and eventually gives rise to a specific cell type. Differentiation can be either stochastic or, when appropriate signals are present, it can be driven to take a specific route. Induced pluripotency has also been recently obtained by overexpressing some genes in a differentiated cell. Here we show that a stochastic dynamical model of genetic networks can satisfactorily describe all these important features of differentiation, and others. The model is based on the emergent properties of generic genetic networks, it does not refer to specific control circuits and it can therefore hold for a wide class of lineages. The model points to a peculiar role of cellular noise in differentiation, which has never been hypothesized so far, and leads to non trivial predictions which could be subject to experimental testing.
  • Seminar: Algorithms for Sequence Finding and Selection Problems
Prof. D.T. Lee, Institute of Information Science & Research Center for IT Innovation, Academia Sinica, Taiwan
Fellow of IEEE, Fellow of ACM, Member of Academia Sinica, Member of Academy of Sciences for the Developing World
Thursday, October 7th, h. 11:00, Sala Seminari DISCo U14, first floor
We consider sequence manipulation problems motivated by problems concerning GC content and GC ratio of DNA sequences in bioinformatics. In this talk we will present algorithms for solving a few problems related to sequence manipulation, including searching subsequences of maximum density and selecting subsequences of a certain density, of a given rank, with or without length restrictions. Problem transformation and utilization of efficient data structures or problem-solving methods will be presented. The problem-solving methods are fundamental to computational problems, which arise, for example, in bioinformatics and in geometric computing.
(Joint work with Dr. Tien-Ching Lin, Institute of Information Science, Academia Sinica, Taiwan.)
  • Seminar: Analisi statistica e bioinformatica di genomi animali
Dr. Alessandra Stella, CeRSA, Parco Tecnologico Padano, Lodi, Italy
  • Seminar: Next-Generation Sequencing: Challenges in Bioinformatics
Dr. Raffaella Rizzi, DISCo, University of Milano-Bicocca, Italy
  • Seminar: Automation of bioinformatics data analysis through workflow management systems
Dr. Paolo Romano, Gruppo di Bioinformatica, Istituto Nazionale per la Ricerca sul Cancro (IST), Genoa, Italy
  • Seminar: microRNA Target Gene Prediction
Prof. Francesco Masulli, Faculty of Sciences of the University of Genova, Italy
  • Seminar: Approcci proteomici allo studio del tumore renale
Prof. Fulvio Magni, Dipartimento di Medicina Sperimentale, Facoltà di Medicina e Chirurgia, Università Degli Studi di Milano-Bicocca, Italy
  • Seminar: Misure di Bloat, Overfitting e Complessità Funzionale in Programmazione Genetica
Dott. Mauro Castelli, DISCo, University of Milano-Bicocca, Italy
  • Seminar: Bioinformatics for mass spectrometry-based proteomics analysis
Dr. Pierluigi Mauri, Istituto di Tecnologie Biomediche, CNR, Milano

Past Events (2009)

  • Seminar: Membrane Systems Combining Variable Molecular Structures with Discretised Reaction Kinetics: From a Toy to a Tool in Systems Biology
Dr. Thomas Hinze, Department of Bioinformatics, School of Biology and Pharmacy, Friedrich-Schiller Universität Jena, Germania
  • Seminar: Conserved coexpression networks: a tool for functional genomics and disease gene prediction
Dott. Paolo Provero, Department of Genetics, Biology and Biochemistry, dell'Università di Torino.
held in honor of Giancarlo Mauri, on the occasion of his 60th birthday.
  • Seminar: Nanorobots in Medicine
Dott. Gianfranco Cerofolini, Dipartimento di Scienze dei Materiali, Università degli Studi di Milano-Bicocca, Milan, Italy
  • Seminar: Combining Structural Bioinformatics and Molecular Modelling to gain insight into protein function
Prof. Laura Bonati, Dipartimento di Scienze dell'Ambiente e del Territorio, Università degli Studi di Milano-Bicocca, Milan, Italy
  • Seminar: High Throughput Bioinformatics Challenges
Dott. Luciano Milanesi, Istituto di Tecnologie Biomediche, CNR, Milan, Italy
  • Seminar: Parameter Estimation for Stochastic Simulation: Applications to Bacterial Chemotaxis
Dr. Dario Pescini, Dipartimento di Informatica Sistemistica e Comunicazione, Università di Milano Bicocca
  • Seminar: Biclustering Algorithms for Biological Data Analysis
Dott. Ilaria Giordani, Dipartimento di Informatica Sistemistica e Comunicazione, Università di Milano Bicocca
  • Seminar: Stochastic Simulations with MESORD. Applications to Small Genetic Networks
Dott. Stefano De Pretis, Laboratorio di Modellistica Molecolare Dipartimento di Biotecnologie e Bioscienze Università di Milano Bicocca

Past Events (2008)

  • Seminar: Knowledge-based bioinformatics for large-scale biological research
Prof. Alberto Riva, University of Florida, Gainsville, FL, USA
A key mechanism for expanding transcriptome and proteome complexity
October 3, 2008
Università di Milano-Bicocca, Milano - Dipartimento di Informatica Sistemistica e Comunicazione
  • Seminar: Systems based on interactions of inhibitions and facilitations
September 5, 2008
Prof. Grzegorz Rozenberg, Leiden Institute of Advanced Computer Science, Leiden University, the Netherlands
Friday September 19, 2008
Romeo Rizzi, Università degli Studi di Udine
Wednesday July 16, 2008
Riccardo Dondi, Università degli Studi di Bergamo
Wednesday July 16, 2008
Anthonath Roslin Sagaya Mary, Università degli Studi di Milano-Bicocca
Wednesday April 9, 2008
Prof. Marco Antoniotti, Università degli Studi di Milano-Bicocca