High Performance Computing and Networking Institute (ICAR)

Description of organizationlogoicar fedele originale fondo bianco
High Performance Computing and Networking Institute (ICAR) is a computer science research institute of Italian National Research Council situated in the south of Italy, with branches in Naples, Cosenza and Palermo. Established in 2002, ICAR activities focus on research, technology transfer, and higher education in the areas of high performance computing and systems (grid and cloud computing, parallel and distributed systems, advanced Internet technology and environments, scientific computing) and intelligent systems with complex functionality (management of very large data source and streams, knowledge representation and knowledge discovery, robotic perceptual systems, intelligent multi-agent systems, multimedia systems). Strategic areas are data analysis, cognitive systems, scientific computing, bioinformatics, sensor networks, and e-health.

Previous experience
Since 2007, the research unit “Integrated Support Systems for Omics Sciences” (ISSOS) at ICAR-CNR, coordinated by Dr. Mario Rosario Guarracino, delivers high performance solutions for high through-put data analysis, image visualization and processing, and data security. Results have been published in top scientific journals, and have advanced insights on gene regulation, proteins interaction, rare genetic disease, cancer cell death detection, security and privacy of patient data, to name a few. Our expertise can enhance activities related to registration, segmentation and interactive 3D visualization of biomedical images; processing, with an holistic view, of patient data using relevant mathematical models; data mining and machine learning methods for diagnostic and prognostic predictions, and drug response; algorithms for both compression and secrecy of patient data. From a software implementation point of view, we can provide solutions for multi-many processor architectures, computer systems close to the data-generating device and high performance computing software infrastructure planning and deployment. Major ongoing projects regard planning and design of software tools for the analysis of omics data with robust mining algorithms for distributed computing environments, the development of integrated informatics tools for genomics, proteomics and transcriptomics, the evaluation of gene variants in the study of inherited disease through large scale analysis of genomic sequences and the evaluation of the effect of specific genes and molecules on transcriptional patterns by means of array hybridization and/or large scale analysis of transcribed sequences.

Profile of staff members
Mario R. Guarracino received a PhD in Mathematics and an Ms in Applied Mathematics. His postdoctoral training from National Research Council focused on low cost high performance architectures for scientific computing. His research interests include high performance scientific computing, machine learning, data mining and bioinformatics.

Diego Romano is researcher on parallel and distributed computing, focusing on Parallel Computer Graphics and Visualization, and software performance evaluation on Multi-Many processor hybrid computing systems.

Lucia Maddalena received the Laurea degree cum laude in mathematics in 1990 and the Ph.D in Applied Mathematics and Computer Science in 1995. Initial researches dealt with parallel computing algorithms, methodologies and techniques, and their application to computer graphics. Subsequent researches are devoted to methods, algorithms and software for image processing and multimedia systems in high-performance computational environments, with application to real-world problems, including digital film restoration, video surveillance, and analysis of bioinformatics images.

Pasqua D’Ambra is senior researcher. She received a PhD in Applied Mathematics and Computer Science and an Ms in Mathematics. Her research interests are focused on high-performance scientific computing with main activities in parallel numerical linear algebra and parallel numerical solution of partial/ordinary differential equations with applications to fluid dynamics and image processing.

Giovanni Schmid received his Laurea degree (cum laude) in Mathematics and his Ph.D. in Applied Mathematics and Computer Science. He is a Certified Information Systems Security Professional (CISSP) by the International Information Security System Certification Consortium (ISC)^2. His research interests are focused on cryptographic protocols for low-power and resource constrained devices, trust models for distributed systems, and privacy-preserving management of sensitive data.

Maria Brigida Ferraro received a PhD in Methodological Statistics and an Ms cum laude in Statistics. Her postdoctoral training from Department of Statistical Sciences Sapienza University of Rome focused on inferential procedures for regression models with random fuzzy sets and fuzzy clustering. She has been collaborating with University of Oviedo since 2007. Her research interests include models and methods for information and uncertainty management in data analysis for classification and forecasting.


Five recent publications relevant to the project

V. Belcastro, F. Gregoretti, V. Siciliano, M. Santoro, G. D'Angelo, G. Oliva, D. di Bernardo: Reverse Engineering and Analysis of Genome-Wide Gene Regulatory Networks from Gene Expression Profiles Using High-Performance Computing. IEEE/ACM Trans. Comput. Biology Bioinform. 9(3): 668-678 (2012)

M.R. Guarracino, P. Xanthopoulos, G. Pyrgiotakis, V. Tomaino, B.M. Moudgil and P.M. Pardalos, Classification of cancer cell death with spectral dimensionality reduction and generalized eigenvalues, Artificial Intelligence in Medicine, vol. 53, pp. 119-125, 2011.

G. Andreotti, M.R. Guarracino, M. Cammisa, A. Correra, and M. V. Cubellis. Prediction of the responsiveness to pharmacological chaperones: lysosomal human alpha-galactosidase, a case of study. Orphanet Journal of Rare Diseases 2010, 5:36.

F. Gregoretti, V. Belcastro, D. di Bernardo, G. Oliva (2010) A Parallel Implementation of the Network Identification by Multiple Regression (NIR) Algorithm to Reverse-Engineer Regulatory Gene Networks. PLoS ONE 5(4): e10179

P. D’Ambra, D. di Serafino, S. Filippone, MLD2P4: a Package of Parallel Algebraic Multilevel Domain Decomposition Preconditioners in Fortran 95, ACM Transactions on Mathematical Software, Vol. 37, N. 3, 2010.