FR | EN

PreBio - Development of a high-performance reconfigurable computing facility for bioinformatics


In the last years, bioinformatics and, more generally speaking, computational biology have crossed the borders of universities and basic-research institutes to enter in the palette of tools used in pharmaceutical and biotechnological industries. Such a migration implies also that bioinformatics and computational biology (BCB) tools should respond to industrial requirements such as high-throughput data processing, standardized performance, cost efficacy, and reusability. The recently approved FP7-EU project PharMEA (Multi-Electrode Array technology-based platform for industrial pharmacology and toxicology drug screening), in which REDS participates actively, constitutes a good example of this new tendency towards industrialized BCB data processing and management. At the same time, there is an increasing interest for developing high-performance reconfigurable computing (HPRC) solutions, applied in particular to accelerate scientific computing tasks offering a very competitive performance-to-cost ratio (usually measured in mega-operations per watt). Besides the intended computational speed-up, other advantages of HPRC with respect to processor-based solutions are lower power consumption for similar computational performance and less space required. The main current drawback is that some hardware-specific programming skills are required to implement algorithms on HPRC platforms. To palliate this, a number of solutions and initiatives are emerging to facilitate porting algorithms from common programming environments (e.g., C language or Matlab) to HPRC platforms. A good example is the HES-SO project Math2Mat, which REDS participates to. The goal of this project is to develop a tool for automatically translating Matlab code into a hardware description language.

PreBio is a self-contained project conceived as the first stage of a longer-term research goal, the HiperBio facility, which will be a HPRC set of tools oriented specifically towards the development of algorithms and solutions for BCB applications. With HiperBio-1 we will explore and establish the technological and methodological bases of HiperBio. To do this, we will use a top-down approach based on the implementation of two selected applications that will allow us to identify and implement a number of basic functional blocks. Goals 1.To establish a basic computational facility for rapid-prototyping and implementation of algorithms for bioinformatics and computational biology (BCB). 2.To implement two selected, industrially relevant, BCB solutions as a proof of the concept. 3.To constitute, based on the developed solutions, a set of hardware-oriented computational “blocks” that will serve as the basic bricks for implementing further advanced BCB solutions. 4.Establish and measure a number of criteria allowing to evaluate a computer-aided hardware development environment and to compare it against the traditional, human-driven, design flow.




REDS
Reconfigurable & embedded Digital Systems Institut REDS
Route de Cheseaux 1
CH-1401 Yverdon-les-Bains
Work+41 24 55 76 330

© 2009 REDS / HEIG-VD