New Algorithmic Approaches to Macromolecular Crystallographic Analysis

Supervisors: Prof. Henry Chapman (DESY, UHH), Prof. Matthias Mnich (TUHH)

The discovery of the double-helix structure of DNA by Franklin, Watson and Crick, and the understanding of the structures of the proteins created from the genetic code are milestones of science in the 20th century. A key technique for these discoveries was X-ray crystallography, which is a Nobel Prize-winning method to analyze the atomic structure within a crystal. The major challenge in obtaining the crystal structure is that the information from the measured diffraction is in most cases insufficient to encode the electron density. Current methods often fit a related structure to the data, leading to inaccuracies.
In this project we will discover macromolecular structures from the full time-dependent dataset, which may be a much more robust way to refine the ensemble of structures and their time-varying populations from the complete data series. You will have the chance to develop, apply, and engineer algorithms which explore and analyze these big data sets, and you will thereby significantly advance our understanding of macromolecular structures. You will use and enhance a variety of mathematical and data-analytic techniques for the design and analysis of your algorithms, such as artificial intelligence, mathematical programming, machine learning, approximation algorithms, and distributed computing.