COMPUTATIONAL PREDICTION OF BETA-SHEET ARRANGEMENT

点击翻译按钮获取中文摘要

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. It is widely believed that the transformation from local ?-helix structures to intermolecular beta-sheets of some human proteins is the main cause of neurodegenerative diseases including some of the most feared and costly diseases such as Alzheimers disease, Parkinsons disease, huntingtons disease, transmissible spongiform encephalopathies (TSEs). However, the exact mechanism of conformation transformation is not clear yet due to the factor the pathogenic products of the process are noncrystalline and insoluble, and therefore it is not possible to use x-ray crystallography and solution NMR to determine the structures. Computational prediction of the transformation may present a plausible approach in this very active neuroscience research area. Unlike alpha-helices, which involve only local amino acids, the beta-sheets may involve long-range interactions between participating strands that may not be necessarily successive in the amino acid sequences. Thus it is difficult to predict sheet arrangement and consequently the predictions have had very limited success. However, these long-range interactions provide information about the topology adopted by a protein sequence. The prediction of arrangement of alpha-strands to form beta-sheet may well be an essential step to predict the 3-D structure of proteins from amino acid sequences. A better understanding of the arrangement of beta-strands to form beta-sheets will not only provide possible solutions to prevent intermolecular beta-sheet formation that causes these neurodegenerative diseases but also contribute to the success of 3-D structure prediction. This project addresses this issue by developing novel computational models to predict the likelihood of two beta-strands to exist adjacently in a beta-sheet and the alignment of the product. Specific aims include: 1) The development of a database of long-range interactions in protein beta-sheets; 2) The development of new computational models based on the support vector machine (SVM) algorithm to predict the likelihood of two beta-strands to exist adjacently in a beta-sheet; 3) The development of new computational models to predict the alignment of two beta-strands if they are predicted to interact with each other as stated in aim 2. Long-term objectives: To better understand inter- and intra- molecular long-range interactions. To apply gained knowledge in neuroscience research, finding solutions for neurodegenerative diseases, and predicting protein 3-D structure.