ABSTRACT
The clinical efficacy and safety of a drug is determined by its molecular properties and targets in humans. However, proteome-wide evaluation of all compounds in humans, or even animal models, is challenging. In this study, we present an unsupervised pretraining deep learning framework, named ImageMol, pretrained on 10 million unlabelled drug-like, bioactive molecules, to predict molecular targets of candidate compounds. The ImageMol framework is designed to pretrain chemical representations from unlabelled molecular images on the basis of local and global structural characteristics of molecules from pixels. We demonstrate high performance of ImageMol in evaluation of molecular properties (that is, the drug's metabolism, brain penetration and toxicity) and molecular target profiles (that is, beta-secretase enzyme and kinases) across 51 benchmark datasets. ImageMol shows high accuracy in identifying anti-SARS-CoV-2 molecules across 13 high-throughput experimental datasets from the National Center for Advancing Translational Sciences. Via ImageMol, we identified candidate clinical 3C-like protease inhibitors for potential treatment of COVID-19. Predicting the properties of a molecule from its structure with high accuracy is a crucial problem in digital drug design. Instead of sequence features, Zeng and colleagues use an image representation of a large collection of bioactive molecules to pretrain a model that can be fine-tuned on specific property prediction tasks.
ABSTRACT
Objective: To analyze the epidemiological history, clinical manifestations, treatment and the short-term prognosis of 31 cases of 2019 novel coronavirus (2019-nCoV) infection in children from six provinces (autonomous region) in northern China. Methods: A retrospective analysis of the epidemiological history, clinical symptoms, signs, laboratory examinations, chest imaging, treatment and the short-term prognosis of 31 cases of 2019-nCoV was conducted. The patients were diagnosed between January 25th, 2020 and February 21st, 2020 in 21 hospitals in 17 cities of six provinces (autonomous region) of Shaanxi, Gansu, Ningxia, Hebei, Henan and Shandong. Results: The age of the 31 children with 2019-nCoV infection was 7 years and 1 month (6 months-17 years). Nine cases (29%) were imported cases. Other 21 cases (68%) had contact with confirmed infected adults. One case (3%) had contact with asymptomatic returnees from Wuhan. Among the 31 children, 28 patients (90%) were family cluster cases. The clinical types were asymptomatic type in 4 cases (13%), mild type in 13 cases (42%), and common type in 14 cases (45%). No severe or critical type existed. The most common symptom was fever (n=20, 65%), including 1 case of high fever, 9 cases of moderate fever, 10 cases of low fever. Fever lasted from 1 day to 9 days. The fever of fifteen cases lasted for ≤3 d, while in other 5 cases lasted >3 d. Other symptoms included cough (n=14, 45%), fatigue (n=3, 10%) and diarrhea (n=3, 10%). Pharyngalgia, runny nose, dizziness, headache and vomiting were rare. In the early stage, the total leukocytes count in peripheral blood decreased in 2 cases (6%), the lymphocytes count decreased in 2 cases (6%), and the platelet count increased in 2 cases (6%).Elevation of C-reactive protein (10%, 3/30), erythrocyte sedimentation rate (19%, 4/21), procalcitonin (4%,1/28), liver enzyme (22%, 6/27) and muscle enzyme (15%, 4/27) occurred in different proportions. Renal function and blood glucose were normal. There were abnormal chest CT changes in 14 cases, including 9 cases with patchy ground glass opacities and nodules, mostly located in the lower lobe of both lungs near the pleural area. After receiving supportive treatment, the viral nucleic acid turned negative in 25 cases within 7-23 days. Among them, 24 children (77%) recovered and were discharged from hospital. No death occurred. Conclusions: In this case series, 2019-nCoV infection in children from six provinces (autonomous region) in northern China are mainly caused by close family contact. Clinical types are asymptomatic, mild and common types. Clinical manifestations and laboratory examination results are nonspecific. Close contact history of epidemiology, nucleic acid detection and chest imaging are important bases for diagnosis of 2019-nCoV infection. After general treatment, the short-term prognosis is good.