prev
next
home
The classified dataset can be used for two search applications. First,
it can be searched with an interface of a newly determined complex to
classify it and obtain hint for its function. Second, it can be
searched with a complete protein. Here, for each representative of the
created clusters we search the query protein for a region similar to
either side of the interface. The binding partner of the selected
region can potentially bind to the protein of interest. Applying the
transformation of the alignment to the binding partner provides a
prediction of its binding mode as would have been done by a docking
algorithm. In this presentation I will focus on the first search type,
which requires a simultaneous alignment between two sides of an
interface and can not be performed by other methods.