For improving low R&D efficiency in drug discovery and development, digital transformation (DX) based on big-data and data science technologies, such as machine learning, artificial intelligence and modeling & simulation, has been implemented in pharmaceutical and biotech industries. The process innovation by DX, called as data-driven drug discovery and development (D5), is supporting the rational decision making from exploratory research to late-stage clinical development, and enables identification of novel target molecules, optimization of lead compounds, clarification of pharmacological or toxicological mechanism-of-action, prioritization of development indications, facilitating of translational research, optimization of clinical trials, and others. In this presentation, I will explain the effectiveness and the technical issues of D5 approach by showing some case-studies related to computational drug repurposing as a good practice in D5 approach. And also, I will show the current application of deep learning and transfer learning for automation and speed-up of image data analytics in non-clinical pharmacological studies and drug screening. The D5 approach is an innovative research framework for leading to successful drug discovery and development by improving the R&D efficiency and changing of research process.