Integration of an event-driven Timepix3 hybrid pixel detector into a cryo-EM workflow

<strong>Abstract</strong> The development of direct electron detectors has played a key role in low-dose electron microscopy imaging applications. Monolithic active-pixel sensor (MAPS) detectors are currently widely applied for cryogenic electron microscopy (cryo-EM); however, they have best performance at 300~kV, have relatively low read-out speed and only work in imaging mode. Hybrid pixel detectors (HPDs) can operate at any energy, have a higher DQE at lower voltage, have unprecedented high time resolution, and can operate in both imaging and diffraction modes. This could make them well-suited for novel low-dose life-science applications, such as cryo-ptychography, iDPC, and liquid cell imaging. Timepix3 is not frame-based, but truly event-based, and can record individual hits with 1.56~ns time resolution. Here, we present the integration of such a detector into a cryo-EM workflow and demonstrate that it can be used for automated data collection on biological specimens. The performance of the detector in terms of MTF and DQE has been investigated at 200~kV and we studied the effect of deterministic blur. We describe a single-particle analysis structure of \SI{3}{\angstrom} resolution and compare it with Falcon3 data collected under the same microscope. These studies could pave the way toward more efficient low-dose single-particle techniques. <strong>Data description</strong> Data has been split up in several different directories. In general: each directory contains individual READMEs <strong>Flat fields</strong> Collected on both TImepix3 and Falcon3 at 200 kV using a Tecnai Arctica microscope. These data have been used for calculating NPS, ToT correction and gain correction. <strong>Knife edge</strong> Collected on both TImepix3 and Falcon3 at 200 kV using a Tecnai Arctica microscope. These data have been used for calculating MTF. <strong>ToT correction calibration file</strong> This calibration file has been used to correct all raw Timepix3 data. Including micrographs deposited in EMPIAR. <strong>Gain correction</strong> Gain correction files calculated from flat field data for several different image formation methods of the Timepix3. The Python script for calculating the gain has been included. <strong>Software</strong> The software tpx3HitParser, tpx3EventViewer and the MTF-NPS-DQE scripts have listed as related identifiers to this entry.**摘要** 直接电子探测器的发展在低剂量电子显微镜成像应用中发挥了关键作用。目前单片有源像素传感器（Monolithic Active-Pixel Sensor, MAPS）探测器已广泛应用于低温电子显微镜（cryo-EM）领域，但该类探测器仅在300 kV电压下性能最优，读出速度相对较低，且仅支持成像模式工作。混合像素探测器（Hybrid Pixel Detectors, HPDs）可在任意能量下工作，在低电压下具备更高的量子检测效率（Detective Quantum Efficiency, DQE），拥有前所未有的高时间分辨率，同时支持成像与衍射两种模式。这使得其非常适用于新型低剂量生命科学应用，例如低温叠层成像（cryo-ptychography）、积分微分相位衬度成像（iDPC）以及液体池成像。Timepix3探测器并非基于帧式成像，而是真正的基于事件式成像，可记录单个击中事件，时间分辨率达1.56 ns。本文介绍了将此类探测器集成至低温电镜工作流程的方案，并验证了其可用于生物样本的自动化数据采集。我们在200 kV电压下对该探测器的调制传递函数（Modulation Transfer Function, MTF）与量子检测效率性能进行了研究，并分析了确定性模糊的影响。本文还描述了一项分辨率达3 Å的单颗粒分析结构，并将其与同一台显微镜下采集的Falcon3数据进行了对比。此类研究可为更高效的低剂量单颗粒技术铺平道路。 **数据说明** 本数据集已拆分至多个不同目录中。一般而言，每个目录均包含独立的README文档。 **平场数据** 本数据在200 kV电压下，使用Tecnai Arctica显微镜采集自Timepix3与Falcon3两款探测器。此类数据用于计算噪声功率谱（Noise Power Spectrum, NPS）、飞越时间校正（Time-over-Threshold, ToT校正）及增益校正。 **刃边数据** 本数据在200 kV电压下，使用Tecnai Arctica显微镜采集自Timepix3与Falcon3两款探测器。此类数据用于计算调制传递函数（MTF）。 **ToT校正校准文件** 本校准文件用于校正所有原始Timepix3数据，包括存入EMPIAR数据库的显微图像。 **增益校正文件** 本增益校正文件基于Timepix3的多种成像模式下的平场数据计算得到，同时附带了用于计算增益的Python脚本。 **软件工具** 本数据集条目已关联tpx3HitParser、tpx3EventViewer两款软件及MTF-NPS-DQE计算脚本作为相关标识符。