Developing the Interaction model of the drug and target for Drug Repositioning and QSAR using the Message Passing Neural Network(MPNN) based on GCN for drug and CNN for target, and model upgrading with the Transformer model.

- Implementation of de novo drug generator using the grammar VAE model for molecular optimization. - Implementation to generate new molecules that possess the property of derived gene expression using the grammar model.

- Constructing the automatic de novo drug design platform of the AI model by the generative adversarial network(GAN) based on the Seq2Seq Autoencoder. (patent pending) - Using the above of the automatic de novo drug design platform, developing two modules of vector operations in the high dimensional

- Developing a Deep Learning model with CNN layers of the physicochemical, 2D Fingerprints of molecular structure info., and 3D conformer features to interpret and predict the relationship between structures of small molecules and their bioactivities for protein targets of diseases.

- The implementation of visualization analysis tools for gene structure and protein domain on the basis of SNV (python)