Automated Annotator: Capturing Expert Knowledge for Free.

Deep learning enabled medical image analysis is heavily reliant on expert annotations which is costly. We present a simple yet effective automated annotation pipeline that uses autoencoder based heatmaps to exploit high level information that can be extracted from a histology viewer in an unobtrusive fashion. By predicting heatmaps on unseen images the model effectively acts like a robot annotator. The method is demonstrated in the context of coeliac disease histology images in this initial work, but the approach is task agnostic and may be used for other medical image annotation applications. The results are evaluated by a pathologist and also empirically using a deep network for coeliac disease classification. Initial results using this simple but effective approach are encouraging and merit further investigation, specially considering the possibility of scaling this up to a large number of users.

Association of Gut Microbiota during Early Pregnancy with Risk of Incident Gestational Diabetes Mellitus.

AIMS: We aimed to assess the association between gut bacterial biomarkers during early pregnancy and subsequent risk of gestational diabetes mellitus (GDM) in Chinese pregnant women. METHODS: Within the Tongji-Shuangliu Birth Cohort study, we conducted a nested case-control study among 201 incident GDM cases and 201 matched controls. Fecal samples were collected during early pregnancy (at 6-15 weeks), and GDM was diagnosed at 24 to 28 weeks of pregnancy. Community DNA isolated from fecal samples and V3-V4 region of 16S rRNA gene amplicon libraries were sequenced. RESULTS: In GDM cases versus controls, Rothia, Actinomyces, Bifidobacterium, Adlercreutzia, and Coriobacteriaceae and Lachnospiraceae spp. were significantly reduced, while Enterobacteriaceae, Ruminococcaceae spp., and Veillonellaceae were overrepresented. In addition, the abundance of Staphylococcus relative to Clostridium, Roseburia, and Coriobacteriaceae as reference microorganisms were positively correlated with fasting blood glucose, 1-hour and 2-hour postprandial glucose levels. Adding microbial taxa to the base GDM prediction model with conventional risk factors increased the C-statistic significantly (P < 0.001) from 0.69 to 0.75. CONCLUSIONS: Gut microbiota during early pregnancy was associated with subsequent risk of GDM. Several beneficial and commensal gut microorganisms showed inverse relations with incident GDM, while opportunistic pathogenic members were related to higher risk of incident GDM and positively correlated with glucose levels on OGTT.

Systemic immune-inflammation index and incident cardiovascular diseases among middle-aged and elderly Chinese adults: The Dongfeng-Tongji cohort study.

BACKGROUND AND AIMS: Systemic immune-inflammation index (SII) has been recently investigated as a novel inflammatory and prognostic marker. SII may be used as an indicator reflecting the progressive inflammatory process in atherosclerosis, although its link to incident cardiovascular disease (CVD) has not been examined in previous studies. This study aims to prospectively assess the association of SII with incident CVD and its main subtypes in Chinese adults. METHODS: Using data from the Dongfeng-Tongji cohort study, 13,929 middle-aged and older adults with a mean age of 62.56 years (range 35-91 years), who were free of CVD and cancer, were included for analysis. The baseline study was conducted in Shiyan city, Hubei province from 2008 to 2009. The SII was calculated as platelet count (/L) × neutrophil count (/L)/lymphocyte count (/L). Cox regression models were used to examine the associations of SII with incident CVD, including stroke and coronary heart disease (CHD). RESULTS: Over a median 8.28 years (maximum 8.98 years) of follow-up, 3386 total CVD cases, including 801 stroke cases and 2585 total CHD cases, were identified. In multivariable Cox regression analyses, higher levels of log-transformed SII were significantly associated with total stroke (HR 1.224, 95% CI 1.065-1.407) and ischemic stroke (HR 1.234, 95% CI 1.055-1.442). For those participants with the highest quartiles of SII versus the lowest quartiles of SII, the HRs were 1.358 (95% CI 1.112-1.658) for total stroke, 1.302 (95% CI 1.041-1.629) for ischemic stroke, and 1.600 (95% CI 1.029-2.490) for hemorrhagic stroke. CONCLUSIONS: SII may serve as a useful marker to elucidate the role of the interaction of thrombocytosis, inflammation, and immunity in the development of cerebrovascular diseases in the middle-aged and elderly population.

GPU-Accelerated Flexible Molecular Docking.

Virtual screening is a key enabler of computational drug discovery and requires accurate and efficient structure-based molecular docking. In this work, we develop algorithms and software building blocks for molecular docking that can take advantage of graphics processing units (GPUs). Specifically, we focus on MedusaDock, a flexible protein-small molecule docking approach and platform. We accelerate the performance of the coarse docking phase of MedusaDock, as this step constitutes nearly 70% of total running time in typical use-cases. We perform a comprehensive evaluation of the quality and performance with single-GPU and multi-GPU acceleration using a data set of 3875 protein-ligand complexes. The algorithmic ideas, data structure design choices, and performance optimization techniques shed light on GPU acceleration of other structure-based molecular docking software tools.

Guiding Conventional Protein-Ligand Docking Software with Convolutional Neural Networks.

The high-performance computational techniques have brought significant benefits for drug discovery efforts in recent decades. One of the most challenging problems in drug discovery is the protein-ligand binding pose prediction. To predict the most stable structure of the complex, the performance of conventional structure-based molecular docking methods heavily depends on the accuracy of scoring or energy functions (as an approximation of affinity) for each pose of the protein-ligand docking complex to effectively guide the search in an exponentially large solution space. However, due to the heterogeneity of molecular structures, the existing scoring calculation methods are either tailored to a particular data set or fail to exhibit high accuracy. In this paper, we propose a convolutional neural network (CNN)-based model that learns to predict the stability factor of the protein-ligand complex and exhibits the ability of CNNs to improve the existing docking software. Evaluated results on PDBbind data set indicate that our approach reduces the execution time of the traditional docking-based method while improving the accuracy. Our code, experiment scripts, and pretrained models are available at https://github.com/j9650/MedusaNet.

[Study advance of depressive animal models and its application in traditional Chinese medicines].

Depression is a kind of mental disease with main symptoms of low mood and lack of pleasure, which seriously endangers human health. An appropriate depressive animal model is of great significance for the study of depression and new antidepressant drugs, while the suitable selection and matching of experimental animals, modeling methods and evaluation indexes are critical to eva-luate the scientificity and effectiveness of the depressive animal model. The study advance of depressive animal models in the aspects of experimental animal selection, modeling principle and method, characteristics, evaluation indexes and their application in traditional Chinese medicine are summarized through the systematic review of relevant literatures in PubMed, CNKI and other databases. The depressive animal modeling methods utilized in recent studies include stress, glucocorticoid induction, reserpine induction, lipopolysaccharide induction, surgical modeling, gene knockout, joint application modeling methods. Stress method is better to simulate the depressive symptoms of clinical patients, whereas there are some deficiencies, such as long modeling time and large cost. The depressive animal models induced by glucocorticoid, reserpine and lipopolysaccharide have the advantages of short modeling time and good controllability, but with a poor reliability. The pathogenesis of surgical modeling is highly matched with that of clinical depressive patients, whereas it has the defect of long postoperative recovery period. Gene knockout models can be used to study the precise role of specific genes in depression. However, its applicability may be restricted in studies on depression. The joint application modeling method can improve its reliability and accuracy, and attracts more and more attention. This paper provides a reference for the selection of animal models in future studies of pathological mechanism of depression, and screening and evaluation of antidepressant drugs.