Challenges in biomolecular network reconstruction

点击翻译按钮获取中文摘要

Bioinformatics is an interdisciplinary area concerned with how best to apply information technology to explore the huge amounts of data in genomics, to support scientific discovery through data integration, information extraction, and knowledge discovery. Continual advances in biotechnology are providing new information about genomes and mechanisms within cells. The costs are rapidly decreasing such that many scientists can use these instruments to produce data on a wide variety of organisms. This volume and variety of data allows scientists unprecedented insight into the working of cells. While wet lab experiments are essential to validate hypotheses, the life sciences are benefiting from computational data analysis to provide insight that leads to hypotheses, to prioritise hypotheses, and to guide experimental design.r The broad field of biological data analysis is following the trend of systems biology that places an emphasis on network models of the cell processes. This holistic view is a framework for integrating many different views provided by the many different datasets produced by today's instruments. The network model can combine coverage of the major cell processes: metabolism, transport, regulation, and signaling. Furthermore, the models may be structured to reflect the cell compartments and to explicitly capture transport of compounds across compartment membranes by the transmembrane transport proteins. Our own study of bioinformatics methods for the steps of network reconstruction has identified several challenges that are difficult yet seem solvable in our opinion with the data available today. The first challenge is improved coverage of transport across the membranes of the cell and its compartments, along with improved prediction of the compound(s) transported by each transmembrane transport protein. The second challenge is improved reconstruction of the transcriptional regulatory network. Furthermore, these challenges should become easier in the intermediate term as more data on more organisms is produced. Our aim is to integrate our algorithms into an existing open-source package for network reconstruction and to allow large scale distributed computation of the reconstruction process. This project provides benefits at several levels. For computer science, the project may contribute new algorithms for sequences, network construction, data integration, data mining, and knowledge discovery. For bioinformatics, the project will provide improved open-source software. For genomics, the software tools will assist experimentalists design experiments and better elucidate the working of cells. For industry, the tools could be applied to the discovery of enzymes for sustainable bio-based processes, or the bio-engineering of novel micro-organisms.