Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorNgo, Thanh Tung
dc.description.abstractProteins play significant roles in various aspects of the structural and functional organization of the cell. Understanding how proteins interact to form complexes is important for the investigation of various biochemical processes. Along with in vitro experiments, in silico methods have been developed to predict protein-protein interaction for high efficiency. One of the most popular computational approaches is protein docking, which predicts the three-dimensional structure of the complex. However, the current docking algorithms only implicitly present the impact of water, which actually has significant effect on the protein interactions. A new re-ranking algorithm, called IFACEwat [1], was introduced. It implemented interfacial water into the protein interfaces to better predict the interaction between two bodies. This project aimed to improve the scoring function used in IFACEwat. The improved scoring function was expected to provide better ranking results, therefor was able to discriminate correct bounds from false positive ones. Nelder–Mead method was implemented to train the best weight set that gives highest success rate. The improved function was shown to improve the old IFACEwat scoring function by increasing both success rate and true positive rate. Compared to initial rigid docking ZDOCK3.0.2, it also obtained better or equivalently well ranking results. Especially for medium and difficult cases, the improved scoring function achieved better success rate compared to another re-ranking technique ZRANK (96% and 100% comparing to 92% and 95% of ZRANK). The experiment results also showed that water terms have an important contribution in the prediction of protein-protein interaction. Among the three new contribution terms, free energy change plays the most important role in the scoring function, leaving this term out will make the performance drop significantly. After the final year project, new sets of weights for IFACEwat scoring function have been found. The new weight sets were shown to improve old IFACEwat scoring function by increasing both success rate and number of hits (true positive structures). From the experiment results, it can be concluded that including interfacial water helps discriminate better true positive from false positive predictions.en_US
dc.format.extent42 p.en_US
dc.rightsNanyang Technological University
dc.subjectDRNTU::Engineering::Computer science and engineering::Computing methodologies::Artificial intelligenceen_US
dc.titleImpact of interfacial water in protein interfacesen_US
dc.typeFinal Year Project (FYP)en_US
dc.contributor.supervisorKwoh Chee Keongen_US
dc.contributor.schoolSchool of Computer Engineeringen_US
dc.description.degreeBachelor of Engineering (Computer Science)en_US
item.fulltextWith Fulltext-
Appears in Collections:SCSE Student Reports (FYP/IA/PA/PI)
Files in This Item:
File Description SizeFormat 
Final Report.pdf
  Restricted Access
1.3 MBAdobe PDFView/Open

Page view(s)

Updated on May 15, 2021


Updated on May 15, 2021

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.