Setting up a brand-new microscope is always a challenge, especially when it is one of the first electron microscope models produced by a company. With CryoARM300 we were feeling ourselves, test pilots. Together with Marcus Fislage and Hirotoshi Furusho we set up a reliable automated data collection using SerialEM with the invaluable help of David Mastronarde and Jaap Brinks. We thoroughly characterized the performance of the microscope as described in the 2020 IUCrJ paper. The scripts for running our CryoARM with K2/3 detector can be downloaded from GitHub.
More recently, with the participation of Adam Schröfel, we set up coma-corrected data collection which spectacularly improved the data collection throughput reaching up to 10,000 images per day without losing the resolution. We reconstructed mouse heavy chain apoferritin to a resolution of 1.7 Å. Even though I believe the resolution can be further improved, we stopped here and went back to work on structural biology challenges.