|DI TUCCI LORENZO||Cycle: XXXII |
Section: Computer Science and Engineering
Tutor: DANIEL FLORIAN
Advisor: SANTAMBROGIO MARCO DOMENICO Major Research topic
:HUG - Hardware for hUman GenomicsAbstract:
In the coming years, human genome research will likely transform medical practices. The unique genetic profile of an individual and the knowledge of molecular basis of diseases are leading to the development of personalized medicines and therapies, but the exponential growth of available genomic data requires a computational effort that may limit the progress of personalized medicine. Within this context, hardware accelerators have proved to be effective in optimizing the ratio between the performance and the power consumption. FPGAs and ASICs allow a high level of parallelism, at a relative low power consumption. The main problem with such architectures is that they require an expert hardware designer in order to produce a highly parallel system.Furthermore, the process of accelerating an algorithm for such architectures called the hardware design flow, it’s a process that it’s at the same time hard and error-prone.
Within this context, we propose the development of a novel hardware and software integrated system, named HUGenomics.The framework aims at becoming an advanced support for personalized medicine research. Thanks to more efficient algorithms and data integration from different biological sources,HUGenomics aims at simplifying the interpretation of biological information and facilitating genomic research process by means of both computational and data visualization tools.The goal of this research project is to create a tool for research in the field of personalized medicine. The system will be equipped with heterogeneous hardware accelerators posing as a challenge finding a way of mapping the specific algorithm on the most suitable hardware architecture, both from a performance and power consumption point of view.The data visualization step of the system raises the challenge of finding a way of automatically aggregating data from multiple sources and finding the best way of showing the results to the user so that the most relevant details are highlighted. The final system will be cloud-based, exploiting Amazon AWS F1 instances.