Research Article
msiDBN: A Method of Identifying Critical Proteins in Dynamic PPI Networks
Algorithm 1
Multisource integrated deep belief nets (msiDBNs).
Input: nd order of adjacency matrices of dynamic networks | and learning rate ; | Output: Weight matrices , bias for each network | and bias for top common layer ; | (1) Initialize Weight matrices , and | (2) repeat | (3) repeat | (4) pick up the samples of proteins at all time points; | (5) for all hidden units do | (6) compute using (14); | (7) sample from ; | (8) end for | (9) for all in do | (10) compute using (13); | (11) sample from ; | (12) end for | (13) for all hidden units do | (14) compute using (14); | (15) end for | (16) until for all samples in | (17) ; | (18) ; | (19) ; | (20) until all parameters are converged |
|