BACKGROUND Nucleic acid test and antibody assay have been employed in the diagnosis for SARS-CoV-2 infection, but the use of viral antigen for diagnosis has not been successfully developed. Theoretically, viral antigen is the specific marker of the virus and precedes antibody appearance within the infected population. There is a clear need of detection of viral antigen for rapid and early diagnosis.METHODS We included a cohort of 239 participants with suspected SARS-CoV-2 infection from 7 centers for the study. We measured nucleocapsid protein in nasopharyngeal swab samples in parallel with the nucleic acid test. Nucleic acid test was taken as the reference standard, and statistical evaluation was taken in blind. We detected nucleocapsid protein in 20 urine samples in another center, employing nasopharyngeal swab nucleic acid test as reference standard. RESULTSWe developed a fluorescence immunochromatographic assay for detecting nucleocapsid protein of SARS-CoV-2 in nasopharyngeal swab sample and urine within 10 minutes. 100% of nucleocapsid protein positive and negative participants accord with nucleic acid test for same samples. Further, earliest participant after 3 days of fever can be identified by the method. In an additional preliminary study, we detected nucleocapsid protein in urine in 73.6% of diagnosed COVID-19 patients.CONCLUSIONS Those findings indicate that nucleocapsid protein assay is an accurate, rapid, early and simple method for diagnosis of COVID-19. Appearance of nucleocapsid protein in urine coincides our finding of the SARS-CoV-2 invading kidney and might be of diagnostic value.

This study conducts crack identification from real-world images containing complicated disturbance information (cracks, handwriting scripts, and background) inside steel box girders of bridges. Considering the multilevel and multi-scale features of the input images, a modified fusion convolutional neural network architecture is proposed. As input, 350 raw images are taken with a consumer-grade camera and divided into sub-images with resolution of 64 × 64 pixels (67,200 in total). A regular convolutional neural network structure is employed as baseline to demonstrate the accuracy benefits from the proposed fusion convolutional neural network structure. The confusion matrix is defined for prediction performance evaluation on the test set. A total of six additional entire raw images are used to investigate the robustness and feasibility of the proposed approach. A binary conversion process based on the optimal entropy threshold method is applied and closely followed to identify the crack pixels in the sub-images. The effect of the super-resolution inputs on accuracy is investigated. Results show that the trained modified fusion convolutional neural network can automatically detect the cracks, handwriting, and background from the raw images. The recognition errors of the fusion convolutional neural network in both the training and validation processes are smaller than those of the regular convolutional neural network. The super-resolution process hurts the general identification accuracy.

Pharmaceutical cocrystals are multicomponent systems in which at least one component is an active pharmaceutical ingredient and the others are pharmaceutically acceptable ingredients. Cocrystallization of a drug substance with a coformer is a promising and emerging approach to improve the performance of pharmaceuticals, such as solubility, dissolution profile, pharmacokinetics and stability. This review article presents a comprehensive overview of pharmaceutical cocrystals, including preparation methods, physicochemical properties, and applications. Furthermore, some examples of drug cocrystals are highlighted to illustrate the effect of crystal structures on the various aspects of active pharmaceutical ingredients, such as physical stability, chemical stability, mechanical properties, optical properties, bioavailability, sustained release and therapeutic effect. This review will provide guidance for more efficient design and manufacture of pharmaceutical cocrystals with desired physicochemical properties and applications.

Purpose: Cancer antigen–specific T cells are key components in antitumor immune response, yet their identification in the tumor microenvironment remains challenging, as most cancer antigens are unknown. Recent advance in immunology suggests that similar T-cell receptor (TCR) sequences can be clustered to infer shared antigen specificity. This study aims to identify antigen-specific TCRs from the tumor genomics sequencing data. Experimental Design: We used the TRUST (Tcr Repertoire Utilities for Solid Tissue) algorithm to assemble the TCR hypervariable CDR3 regions from 9,700 bulk tumor RNA-sequencing (RNA-seq) samples, and developed a computational method, iSMART, to group similar TCRs into antigen-specific clusters. Integrative analysis on the TCR clusters with multi-omics datasets was performed to profile cancer-associated T cells and to uncover novel cancer antigens. Results: Clustered TCRs are associated with signatures of T-cell activation after antigen encounter. We further elucidated the phenotypes of clustered T cells using single-cell RNA-seq data, which revealed a novel subset of tissue-resident memory T-cell population with elevated metabolic status. An exciting application of the TCR clusters is to identify novel cancer antigens, exemplified by our identification of a candidate cancer/testis gene, HSFX1, through integrated analysis of HLA alleles and genomics data. The target was further validated using vaccination of humanized HLA-A*02:01 mice and ELISpot assay. Finally, we showed that clustered tumor-infiltrating TCRs can differentiate patients with early-stage cancer from healthy donors, using blood TCR repertoire sequencing data, suggesting potential applications in noninvasive cancer detection. Conclusions: Our analysis on the antigen-specific TCR clusters provides a unique resource for alternative antigen discovery and biomarker identification for cancer immunotherapies.