Novel Approach for Drug Discovery using Neural Network Back Propagation Algorithm

Master's Thesis from the year 2017 in the subject Computer Science - Bioinformatics, grade: 90, Punjabi University (Guru Nanak Dev Engineering College, Ludhiana), course: Masters of Technology, language: English, abstract: Computational screening of databases has gained immense popularity in the pharmaceutical research and development. To do such screening tests the Virtual Screening technique is used. It uses computer-based algorithms and methods which takes into consideration a lot of parameters to discover new ligands on the basis of biological structures. The process of discovering new drugs has now become a crucial factor for all the Pharmaceutical Industries. Acceleration of Virtual screening would provide an edge to save the resources as well as the time required. Here, the effectual implementation of the parallel architecture of CUDA and GP-GPU for the acceleration of Virtual screening has been analyzed. The implementation is in CUDA programming models. This implementation tries to take maximum advantage of a GPU to give a better solution in the process of drug discovery. The result which would be higher speedup when implemented on CUDA GPU architecture to provide the better solution considering performance & cost ratio.