Loading...

Learning Molecular Properties Using Deep Learning

Moradi, Parsa | 2018

1757 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 51510 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Hossein Khalaj, Babak
  7. Abstract:
  8. Design and production of a drug is a very time and money consuming process. It takes more than a decade and about 2.5 million dollars on various stages to design a drug. Attempts to reduce this cost and time to market will make drugs available to customers at a more reasonable time. Some stages such as animal testing phase and clinical trials, can not be replaced and must take place in practice. Fortunately, some laboratory steps are interchangeable with software algorithms. These algorithms can significantly reduce the cost and time to market of the drug if they are accurate enough. On the other hand, the remarkable results of machine learning, in particular, Deep Neural Networks, in areas such as image processing and audio processing have led other scholars to come to these networks to solve their research problems. Research has shown that these networks can be useful in other areas if sufficient data is available. In this study, a neural network is presented that has the ability to learn and predict the various properties of chemical molecules. The network input is the graph representation of the molecules and is applicable to every data that is representable by a graph. The structure of the network is such that after the learning process, various visualizations can be made on the model to evaluate how the network was trained. This approach has reached or enhanced the results of the previous methods on a variety of datasets with different issues (regression, binary and multi-class classification)
  9. Keywords:
  10. Molecular Properties Prediction ; Molecular Machine Learning ; Attention Mechanism ; Deep Learning ; Unstructured Data ; Deep Neural Networks

 Digital Object List

 Bookmark

...see more