The National Cryo-EM Facility
FFT, the broad diffraction peak corre- sponding to distances of water mole- cules in vitreous ice, is dependent on ice thickness. Tis can be used as a rough rule of thumb. As can be observed in Figure 3, in ice that is thinner than ∼50 nm, no vitreous ice ring is observed in the power spectrum, whereas with ice thickness >50 nm, the ring becomes progressively more prominent in the power spectrum. Te
thickness of ice can some-
Figure 4: Effect of ice thickness on data quality: (A) Ice thickness distribution for a single NCEF imaging session. (B) CTFFIND4 resolution limit versus ice thickness for each image in the same imaging session. We find that ice thickness is usually a good predictor for resolution limits (as determined by CTF fitting with CTF- FIND4), with only images that have ice thinner than ∼50 nm resulting in achieved resolutions between 3 Å and 4 Å. (C) User-reported resolution values versus average ice thickness of the corresponding imaging session for a subset of the imaging sessions where ice thickness measurements were carried out. (D) User-reported resolution versus average motion between frame 1 and frame 2 of the corresponding imaging session. While ice thickness values show strong correlation with reported resolution, no obvious correlation is observed between average motion and the reported resolution.
average dose (and the known total dose during the session), one must determine Λ on the system at conditions close to standard single-particle imaging conditions. Λ was mea- sured to be 250 nm according to the method described by [5] and [6]. With the measured value Λ, Equation (2) can be used to determine the average ice thickness of each image in an imaging session using the dose per pixel (or per Å2 sured from the image. Te value of Dtot
) mea- is measured carefully
before each data collection and specifically set to a particular value (standard is 40 or 50 e/ Å2
stability on the Titan Krios is excellent, and Dtot change significantly (<1%) during a 2–3 day run.
). In our experience, beam does not
Effect of Ice Thickness on Data Quality Since ice thickness seems to be a very important factor
to determine data quality, it should be strongly considered when optimizing plunging conditions and screening grids. Te visual appearance of the “vitreous ice ring” in the image
2020 May •
www.microscopy-today.com
times vary widely within a single grid, as shown in an example presented in Figure 4A. Te information limit is also determined by ice thickness, with thinner regions producing images with information extending to higher resolu- tion (Figure 4B) [4]. As users have shared results obtained from data collected at NCEF, a measurable trend has emerged that correlates resolution with ice thick- ness. Specifically, thinner ice strongly correlates with higher reported resolu- tions (Figure 4C). Interestingly, there was no obvious correlation between the extent of motion in the movie frames and the reported resolution (Figure 4D), indicating that the soſtware for motion correction available to most users is suf- ficient enough to compensate for beam- induced and stage motion that occurs during data collection.
Summary and Future Perspective Since its inception in May 2017, the NCEF has executed over 350 data col-
lection runs, providing services to 72 investigators in 39 insti- tutions (Figure 5), operating at an overall uptime of >90%. Operations between May 2017 and February 2020 were carried out with a single Titan Krios microscope, but as of April 2020, the facility will operate with two Titan Krios microscopes, each equipped with Falcon 3EC and GIF/K3 cameras. Automated imaging on the K3 cameras is performed using either Latitude S (Gatan, Inc.) or SerialEM soſtware [7], whereas collection on the Falcon 3EC is performed with EPU (Termo Fisher Scien- tific). Te average size of data collected per session increased from under a terabyte to several terabytes as we transitioned from a K2 to K3 camera, but the overall trends in data size (Figure 6) help us anticipate requirements for data storage. As we look ahead to the coming years, we foresee that biologists with minimal cryo-EM or structural biology experience will likely constitute the dominant user community. We therefore anticipate enhancing our services to offer progressively more advanced feedback by providing image processing results so
15
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76 |
Page 77 |
Page 78 |
Page 79 |
Page 80 |
Page 81 |
Page 82 |
Page 83 |
Page 84
