Selected article for: "human protein and protein pair"

Author: Kshirsagar, Meghana; Carbonell, Jaime; Klein-Seetharaman, Judith
Title: Multitask learning for host–pathogen protein interactions
  • Document date: 2013_7_1
  • ID: sdgt2ms5_20
    Snippet: We consider each human-bacteria PPI prediction problem as one task. The predicton problem is posed as a binary classification task, with each instance x i being a pair of proteins 5b, h4, where one protein is the bacterial protein 'b' (e.g. Y. pestis) and the other 'h' is the host protein (i.e. human). The class-label y i 2 fþ1, À 1g represents interacting and noninteracting proteins respectively. Features are defined for every proteinpair usin.....
    Document: We consider each human-bacteria PPI prediction problem as one task. The predicton problem is posed as a binary classification task, with each instance x i being a pair of proteins 5b, h4, where one protein is the bacterial protein 'b' (e.g. Y. pestis) and the other 'h' is the host protein (i.e. human). The class-label y i 2 fþ1, À 1g represents interacting and noninteracting proteins respectively. Features are defined for every proteinpair using various properties of the individual proteins and combining them all into a single feature vector. The positive class in our training data comprises the known human-bacterial PPI which are obtained from databases like PHISTO (Tekir et al., 2012) . The construction of the negativeclass data is explained in Section 5.

    Search related documents:
    Co phrase search for related documents