Visualization Symposium, IEEE Pacific (2012)
Songdo, Korea (South)
Feb. 28, 2012 to Mar. 2, 2012
Chandrajit Bajaj , University of Texas at Austin, USA
Discoveries in computational molecular - cell biology and bioinformatics promise to provide new therapeutic interventions to disease. With the rapid growth of sequence and structural information for thousands of proteins and hundreds of cell types, computational processing are a restricting factor in obtaining quantitative understanding of molecular-cellular function. Processing and analysis is necessary both for input data (often from imaging) and simulation results. To make biological conclusions, this data must be input to and combined with results from computational analysis and simulations. Furthermore, as parallelism is increasingly prevalent, utilizing the available processing power is essential to development of scalable solutions needed for realistic scientific inquiry. However, complex image processing and even simulations performed on large clusters, multi-core CPU, GPU-type parallelization means that naïve cache unaware algorithms may not efficiently utilize available hardware. Future gains thus require improvements to a core suite of algorithms underpinning the data processing, simulation, optimization and visualization needed for scientific discovery. In this talk, I shall highlight current progress on these algorithms as well as provide several challenges for the visualization community.
C. Bajaj, "Quantitative visualization in the computational biological sciences," Visualization Symposium, IEEE Pacific(PACIFICVIS), Songdo, Korea (South), 2012, pp. xi.