ObjectiveTo detect the changes of gastroesophageal reflux disease (GERD) with high resolution endoscopy,and to explore its diagnostic value for nonerosive reflux disease (NERD). Methods From April 2007 to January 2008, consecutive out-patients visiting the Department of Gastroenterology due to continuous or recurrent symptoms of acid reflux, heartburn, cardiac-like chest pain for at least three months and volunteer healthy controls were recruited to the study. The subjects were classified into the normal group ( n =48 ), the NERD group ( n =70), the erosive esophagitis (EE) group ( n =70), and the Barrett esophagus (BE) group ( n =48). All subjects underwent endoscopy, and the shape of Z-lines, the shape of mucosal pits, mucosa roughness above Z-line and the shape of mucosal blood vessels were observed. The NERD changes were analyzed with optimal scale. ResultsZ-line shape of NERD group showed a significant difference from that of EE and BE groups ( P ＜ 0. 01 ), which was not different from that of normal controls ( P ＞0. 01 ). The shape of mucosal pits of NERD group was different from that of EE and BE (P ＜0. 01 ), which also was not different from that of control (P＞0. 01 ). Roughness of the mucosa above Z-line of NERD group was different from other 3 groups (P ＜0. 05). And the shape of mucosal blood vessels of NERD was different from the control (P＜0. 05), but was not from two others. Features of NERD relative images were thin,straight and spiral blood vessels, bar-like pits, plat mucosa, round, smooth and wide tooth-like Z-line and mucosa of white particle hypertrophy. ConclusionHigh resolution endoscopic features of most NERD patients are cloudy, white and rough mucosal surface, with white particular hypertrophy, spiral vascular dilation, extending to Z-line. These features can be indicators of NERD.

OBJECTIVETo screen the HIV-1 entry inhibitors targeting HIV-1 gp120 from the IBS natural product database by virtual screening based on the binding mode of the neutralizing antibody VRC01 with HIV-1 gp120 and investigate the anti-viral activities of the inhibitors and their action mechanisms.

METHODSThe binding interaction of the candidate molecules binding gp120 and changes of the binding free energy were analyzed by MM-PBSA calculation. The anti-HIV-1 activities of the tested compounds were detected by HIV-1 pseudotyped virus, laboratory-adapted HIV-1 and a cell-cell fusion assay. The cytotoxicity of the studied molecules was examined by XTT colorimetric assay. The mechanisms of the anti-viral activities of the candidate molecules were analyzed using enzyme-linked immunosorbent assay.

RESULTSA total of 19 molecules with distinct reduction of the binding free energy after binding with gp120 were screened from 40000 molecules. Among them, NC-2 showed anti-HIV-1 activities against HIV-1 pseudotyped virus and laboratory-adapted HIV-1, and was capable of blocking HIV-1 envelope-mediated cell-cell fusion. The IC50 of NC-2 for inhibiting HIV-1IIIB and pseudotyped HIV-1JRFL infection were 1.95∓0.44 µmol/L and 10.58∓0.13 µmol/L, respectively. The results of ELISA suggested that NC-2 could inhibit the binding of HIV-1 gp120 to CD4 without blocking the formation of gp41 six-helix bundle in vitro.

CONCLUSIONThis computer-based virtual screening method can be used to screen HIV-1 entry inhibitors targeting gp120. Using this virtual screening approach combined with anti-viral activity screening, we obtained a potent HIV-1 entry inhibitor NC-2 with novel structure.

Anti-HIV Agents ; pharmacology ; Antibodies, Monoclonal ; pharmacology ; Antibodies, Neutralizing ; pharmacology ; Binding Sites ; Cell Fusion ; Cell Line ; Drug Discovery ; Drug Evaluation, Preclinical ; HIV Antibodies ; pharmacology ; HIV Envelope Protein gp120 ; antagonists & inhibitors ; HIV-1 ; drug effects ; Humans ; Microbial Sensitivity Tests

Advances in biological and medical technologies have been providing us explosive volumes of biological and physiological data, such as medical images, electroencephalography, genomic and protein sequences. Learning from these data facilitates the understanding of human health and disease. Developed from artificial neural networks, deep learning-based algorithms show great promise in extracting features and learning patterns from complex data. The aim of this paper is to provide an overview of deep learning techniques and some of the state-of-the-art applications in the biomedical field. We first introduce the development of artificial neural network and deep learning. We then describe two main components of deep learning, i.e., deep learning architectures and model optimization. Subsequently, some examples are demonstrated for deep learning applications, including medical image classification, genomic sequence analysis, as well as protein structure classification and prediction. Finally, we offer our perspectives for the future directions in the field of deep learning.
Algorithms ; Computational Biology ; methods ; Diagnostic Imaging ; Genomics ; methods ; Humans ; Image Interpretation, Computer-Assisted ; methods ; Machine Learning ; Neural Networks (Computer) ; Protein Structure, Secondary ; Proteins ; metabolism