FIGURE 6.

(A) Shows a classical deep neural network and its application in protein-ligand binding studies. For a deep neural network, the lower layers usually extract low-level feature, while the higher layers learn the high-level feature. (B) Illustrates the screening of a drug-nanoparticle complex (DNC) by using a deep neural network model and DNC interactions with proteins in the pathway of AD. The neurons are shown in red and green colors interactively. Aβ peptides are marked with blue color whereas gold nanoparticles and inhibitors are indicated as orange and gray color respectively.