Deep Learning Algorithms for Predicting Tandem Mass Spectra of Peptides
Sophie Shelburn
Undergraduate Researcher
Biology Major (College of Arts & Sciences)
Haixu Tang
Faculty Mentor
Haixu Tang (Luddy School of Informatics, Computing and Engineering)
Project Description
Previously, we have developed a deep learning algorithm PredFull for predicting the full tandem mass spectra of peptides without modifications. In this project, we attempt to investigate if the deep learning methods can be extended to spectra prediction for peptides containing post-translation modifications (PTMs). The project will involve the assembly a large training datasets of mass spectra and the training of deep neural networks using these data.
Technology or Computational Component
Development and training of deep neural networks (DNNs) using the platform of Keras/Tensorflow. The student will be work with Ph.D students in my lab to develop these models. The student does not need to have prior experiences in machine learning, but some experience in python programming will be helpful.