Deep Spatio-Temporal Network for Low SNR Cryo-EM Movie Frame Enhancement.

Cryo-EM in single particle analysis is known to have low SNR and requires to utilize several frames of the same particle sample to restore one high-quality image for visualizing that particle. However, the low SNR of cryo-EM movie and motion caused by beam striking make the task very challenging. Video enhancement algorithms in computer vision shed new light on tackling such tasks by utilizing deep neural networks. However, they are designed for natural images with high SNR. Meanwhile, the lack of ground truth in cryo-EM movie seems to be one major limiting factor of the progress. Hence, we present a synthetic cryo-EM movie generation pipeline, which can produce realistic diverse cryo-EM movie datasets with low-SNR movie frames and multiple ground truth values. Then we propose a deep spatio-temporal network (DST-Net) for cryo-EM movie frame enhancement trained on our synthetic data. Spatial and temporal features are first extracted from each frame. Spatio-temporal fusion and high-resolution re-constructor are designed to obtain the enhanced output. For evaluation, we train our model on seven synthetic cryo-EM movie datasets and infer on real cryo-EM data. The experimental results show that DST-Net can achieve better enhancement performance both quantitatively and qualitatively compared with others.

Deep Learning Model for Predicting the Pathological Complete Response to Neoadjuvant Chemoradiotherapy of Locally Advanced Rectal Cancer.

Objective: This study aimed to develop an artificial intelligence model for predicting the pathological complete response (pCR) to neoadjuvant chemoradiotherapy (nCRT) of locally advanced rectal cancer (LARC) using digital pathological images. Background: nCRT followed by total mesorectal excision (TME) is a standard treatment strategy for patients with LARC. Predicting the PCR to nCRT of LARC remine difficulty. Methods: 842 LARC patients treated with standard nCRT from three medical centers were retrospectively recruited and subgrouped into the training, testing and external validation sets. Treatment response was classified as pCR and non-pCR based on the pathological diagnosis after surgery as the ground truth. The hematoxylin & eosin (H&E)-stained biopsy slides were manually annotated and used to develop a deep pathological complete response (DeepPCR) prediction model by deep learning. Results: The proposed DeepPCR model achieved an AUC-ROC of 0.710 (95% CI: 0.595, 0.808) in the testing cohort. Similarly, in the external validation cohort, the DeepPCR model achieved an AUC-ROC of 0.723 (95% CI: 0.591, 0.844). The sensitivity and specificity of the DeepPCR model were 72.6% and 46.9% in the testing set and 72.5% and 62.7% in the external validation cohort, respectively. Multivariate logistic regression analysis showed that the DeepPCR model was an independent predictive factor of nCRT (P=0.008 and P=0.004 for the testing set and external validation set, respectively). Conclusions: The DeepPCR model showed high accuracy in predicting pCR and served as an independent predictive factor for pCR. The model can be used to assist in clinical treatment decision making before surgery.

Early triage of critically ill COVID-19 patients using deep learning.

The sudden deterioration of patients with novel coronavirus disease 2019 (COVID-19) into critical illness is of major concern. It is imperative to identify these patients early. We show that a deep learning-based survival model can predict the risk of COVID-19 patients developing critical illness based on clinical characteristics at admission. We develop this model using a cohort of 1590 patients from 575 medical centers, with internal validation performance of concordance index 0.894 We further validate the model on three separate cohorts from Wuhan, Hubei and Guangdong provinces consisting of 1393 patients with concordance indexes of 0.890, 0.852 and 0.967 respectively. This model is used to create an online calculation tool designed for patient triage at admission to identify patients at risk of severe illness, ensuring that patients at greatest risk of severe illness receive appropriate care as early as possible and allow for effective allocation of health resources.

Particle segmentation algorithm for flexible single particle reconstruction.

As single particle cryo-electron microscopy has evolved to a new era of atomic resolution, sample heterogeneity still imposes a major limit to the resolution of many macromolecular complexes, especially those with continuous conformational flexibility. Here, we describe a particle segmentation algorithm towards solving structures of molecules composed of several parts that are relatively flexible with each other. In this algorithm, the different parts of a target molecule are segmented from raw images according to their alignment information obtained from a preliminary 3D reconstruction and are subjected to single particle processing in an iterative manner. This algorithm was tested on both simulated and experimental data and showed improvement of 3D reconstruction resolution of each segmented part of the molecule than that of the entire molecule.

EMBuilder: A Template Matching-based Automatic Model-building Program for High-resolution Cryo-Electron Microscopy Maps.

The resolution of electron-potential maps in single-particle cryo-electron microscopy (cryoEM) is approaching atomic or near- atomic resolution. However, no program currently exists for de novo cryoEM model building at resolutions exceeding beyond 3.5 Å. Here, we present a program, EMBuilder, based on template matching, to generate cryoEM models at high resolution. The program identifies features in both secondary-structure and Cα stages. In the secondary structure stage, helices and strands are identified with pre-computed templates, and the voxel size of the entire map is then refined to account for microscopic magnification errors. The identified secondary structures are then extended from both ends in the Cα stage via a log-likelihood (LLK) target function, and if possible, the side chains are also assigned. This program can build models of large proteins (~1 MDa) in a reasonable amount of time (~1 day) and thus has the potential to greatly decrease the manual workload required for model building of high-resolution cryoEM maps.

Near-Atomic Resolution Structure Determination of a Cypovirus Capsid and Polymerase Complex Using Cryo-EM at 200kV.

Single-particle cryo-electron microscopy (cryo-EM) allows the high-resolution structural determination of biological assemblies in a near-native environment. However, all high-resolution (better than 3.5Å) cryo-EM structures reported to date were obtained by using 300kV transmission electron microscopes (TEMs). We report here the structures of a cypovirus capsid of 750-Å diameter at 3.3-Å resolution and of RNA-dependent RNA polymerase (RdRp) complexes within the capsid at 3.9-Å resolution using a 200-kV TEM. The newly resolved structure revealed conformational changes of two subdomains in the RdRp. These conformational changes, which were involved in RdRp's switch from non-transcribing to transcribing mode, suggest that the RdRp may facilitate the unwinding of genomic double-stranded RNA. The possibility of 3-Å resolution structural determinations for biological assemblies of relatively small sizes using cryo-EM at 200kV was discussed.