High-performance computing, or HPC, is the answer to the high volumes of information available to solve the complex problems currently faced by organizations and institutions. High-performance computers, such as the one at the Center for Bioinformatics and Computational Biology of Colombia BIOS, combine high capacities and performance with highly qualified human talent. In the era of Big Data, a large infrastructure for data processing is essential since it represents not only an ally for research in Colombia but also a strategic asset of national interest.
The visualization wall of scientific information is one of the most advanced scientific research tools in Latin America. It spans 32 full HD screens that altogether provide a resolution of 66.3 megapixels and can be controlled through three different ways:
- A Tablet or Smartphone
- Kinect-type devices
- Emotiv EEG interface
This project was possible thanks to financial resources provided by the royalties project Caldas Bio-region, which was executed by the Government of Caldas and operated by BIOS. Our research team at BIOS is working on man-machine interface applications, analyses of medical diagnostic images, and molecular docking processes. Besides these applications, the visualization wall provides a broad range of uses for various applied fields. It can be operated in two ways:
- Through a video server for collaborative work that allows to distribute various image sources throughout the entire visualization wall.
- Through a direct connection to graphical processing nodes, which allows to perform real-time high-complexity processing and visualization (work in progress)
Within its own server, BIOS has implemented a version of the popular genome sequence visualizer Jbrowse. This tool allows the users the analysis of annotation results in a more dynamic fashion through a website.
At the present time, large amounts of data are currently being generated through new sequencing technologies and experiments that include gene and protein analysis. However, a considerable amount of this data remains without being characterized, even though the correct annotation of enzymatic functions is critical for the reconstruction of metabolic networks. ZYFCON (en ZYme Function Prediction COmbining Networks) is a tool designed to predict enzymatic functions in sets of coding genes, based upon complete genome sequence data.