Data Reduction and Analysis for Continuous-Tilt Electron Diffraction

Microcrystal electron diffraction MicroED) is a powerful technique promising to fill a longstanding need for rapid and accurate high-resolution structural characterization of large molecules. MicroED can acquire data using crystals that are orders of magnitude smaller than X-ray crystallography and doesn’t require access to synchrotron radiation beamlines. Despite the widespread excitement over MicroED, its actual use has grown slowly because of the need for specialized equipment, insufficient sensitivity of that equipment, and difficulties handling the large quantities of raw data produced. We propose to develop a high-speed, high-sensitivity method for MicroED that provides higher-quality data than present systems and potentially over two million times data compression using currently available direct detection cameras to make MicroED widely available to chemists and structural biologists. Using a high framerate direct electron camera to record continuous tilt diffraction data sampling in 0.04-degree steps at 100 fps, the system will collect 2500 frames in 25 seconds. We will reduce the data by parameterizing each diffraction spot observed in the “movie” achieving very high data compression. Additionally, the parameterization will improve data quality by using a least-squares method over many observations. During Phase I, we will determine the parameters necessary to describe each diffraction spot in a pattern with sufficient accuracy to result in a high-quality reconstruction. We will then implement software to automatically reduce a large diffraction tilt series into a list of diffraction spot parameters. Finally, we will compare the new parameterized method to traditional methods using the entire raw data set and optimize our technique until we have a robust parameterization method. Prior to Phase I we have demonstrated high- speed collection of a continuous tilt series using our state-of-the art DDD® camera system and have achieved a sub-angstrom reconstruction using traditional methods. In Phase II we plan to develop automated electron diffraction collection, processing and 3D reconstruction software. We fully expect MicroED to become an indispensable tool for synthetic organic chemistry, materials science, structural biology, and pharmaceutical development. Offering superior resolution on smaller specimens with less expensive equipment and in a shorter period of time than X-ray diffraction, the technique is poised to become the next NMR or Mass Spectrometry.