Research Summary

Research Area

My current research area is about computational biology using deep learning and AI. We study and apply machine learning algorithms to solve problems in this area. I am also working on classify bird images using computer vision and deep learning technology. The topic I am working on includes:

  • Protein Quality Assessment
  • Protein Tertiary Structure Refinement
  • Protein Tertiary Structure Prediction
  • Machine Learning and Deep Learning on computational biology
  • Computer Vision

Projects

MUFOLD2 – QA & Refinement

MUFOLD is a comprehensive platform to do protein tertiary structure prediction. MUFOLD2 is an objected oriented version of original MUFOLD. We are going to provide a platform for computational protein structure prediction with efficiency and accuracy to help experimental biologists understand structures and functions of the proteins of their interest thereby facilitating hypotheses for experimental design.

  • MUFOLD2 Quality Assessment evaluates the quality of decoys and do model – selection based on machine learning techniques.
  • MUFOLD2 Refinement try to refinement the weak part of a given protein tertiary structure to improve its performance.

Bird recognition and counting from aerial images

Developed machine learning system using deep learning and image process to detect and then count the number of birds from aerial images. The system developed in the project will figure out how to detect small object in the image using computer vision and image process technique.

Publications

  • Junlin Wang, Zhaoyu Li, and Yi Shang. “New Deep Neural Networks for Protein Model Evaluation.” 2017 IEEE 29th International Conference on Tools with Artificial Intelligence (ICTAI). IEEE, 2017.
  • Wenbo Wang, Junlin Wang, Dong Xu, and Yi Shang, “Two New Heuristic Methods for Protein Model Quality Assessment,” IEEE/ACM Transactions on Computational Biology and Bioinformatics, in press.
  • Wenbo Wang, Zhaoyu Li, Junlin Wang, Dong Xu and Yi Shang, “PSICA: a fast and accurate web-based platform for protein model quality analysis,”, Nuclear acid research web server issue, 2019.
  • Junlin Wang. Machine learning methods for evaluating the quality of a single protein model using energy and structural properties. Diss. University of Missouri–Columbia, 2015.