Skip to main content
. 2022 May 19;44(5):2287–2299. doi: 10.3390/cimb44050155

Figure 2.

Figure 2

Architecture of our proposed model DeepMHADTA. The model combines sequence and structure information of protein and drug via CNN block for feature extraction, and the embeddings are concatenated as input to fully connected layer for quantitative binding affinity prediction.