An Unsupervised Text Mining Method for Relation Extraction from Biomedical Literature

The wealth of interaction information provided in biomedical articles motivated the implementation of text mining approaches to automatically extract biomedical relations. This paper presents an unsupervised method based on pattern clustering and sentence parsing to deal with biomedical relation extraction. Pattern clustering algorithm is based on Polynomial Kernel method, which identifies interaction words from unlabeled data; these interaction words are then used in relation extraction between entity pairs. Dependency parsing and phrase structure parsing are combined for relation extraction. Based on the semi-supervised KNN algorithm, we extend the proposed unsupervised approach to a semi-supervised approach by combining pattern clustering, dependency parsing and phrase structure parsing rules. We evaluated the approaches on two different tasks: (1) Protein–protein interactions extraction, and (2) Gene–suicide association extraction. The evaluation of task (1) on the benchmark dataset (AImed corpus) showed that our proposed unsupervised approach outperformed three supervised methods. The three supervised methods are rule based, SVM based, and Kernel based separately. The proposed semi-supervised approach is superior to the existing semi-supervised methods. The evaluation on gene–suicide association extraction on a smaller dataset from Genetic Association Database and a larger dataset from publicly available PubMed showed that the proposed unsupervised and semi-supervised methods achieved much higher F-scores than co-occurrence based method.     

BP神经网络优化槐花中芦丁的提取工艺

为提高槐花加工过程中芦丁的得率,以槐花为原料提取芦丁,采用正交试验设计收集试验数据,利用BP神经网络的自学习能力,通过仿真和评估,优化其提取工艺参数。结果表明:BP神经网络与正交实验方法相结合,能更好的利用已知信息,最佳仿真提取条件为:pH 9.50,乙醇体积分数60%,时间1.45 h,料液比1∶21,运用该提取条件用于实际研究中,测得实际得率为13.51%,试验结果优于正交试验,获得较为理想的目标值。     

A Computationally Efficient Correlational Neural Network for Automated Prediction of Chronic Kidney Disease

ObjectivesIn this paper, we propose a computationally efficient Correlational Neural Network (CorrNN) learning model and an automated diagnosis system for detecting Chronic Kidney Disease (CKD). A Support Vector Machine (SVM) classifier is integrated with the CorrNN model for improving the prediction accuracy.Material and methodsThe proposed hybrid model is trained and tested with a novel sensing module. We have monitored the concentration of urea in the saliva sample to detect the disease. Experiments are carried out to test the model with real-time samples and to compare its performance with conventional Convolutional Neural Network (CNN) and other traditional data classification methods.ResultsThe proposed method outperforms the conventional methods in terms of computational speed and prediction accuracy. The CorrNN-SVM combined network achieved a prediction accuracy of 98.67%. The experimental evaluations show a reduction in overall computation time of about 9.85% compared to the conventional CNN algorithm.ConclusionThe use of the SVM classifier has improved the capability of the network to make predictions more accurately. The proposed framework substantially advances the current methodology, and it provides more precise results compared to other data classification methods.     

Transmission Risks of Schistosomiasis Japonica: Extraction from Back-propagation Artificial Neural Network and Logistic Regression Model

Background

The transmission of schistosomiasis japonica in a local setting is still poorly understood in the lake regions of the People''s Republic of China (P. R. China), and its transmission patterns are closely related to human, social and economic factors.

Methodology/Principal Findings

We aimed to apply the integrated approach of artificial neural network (ANN) and logistic regression model in assessment of transmission risks of Schistosoma japonicum with epidemiological data collected from 2339 villagers from 1247 households in six villages of Jiangling County, P.R. China. By using the back-propagation (BP) of the ANN model, 16 factors out of 27 factors were screened, and the top five factors ranked by the absolute value of mean impact value (MIV) were mainly related to human behavior, i.e. integration of water contact history and infection history, family with past infection, history of water contact, infection history, and infection times. The top five factors screened by the logistic regression model were mainly related to the social economics, i.e. village level, economic conditions of family, age group, education level, and infection times. The risk of human infection with S. japonicum is higher in the population who are at age 15 or younger, or with lower education, or with the higher infection rate of the village, or with poor family, and in the population with more than one time to be infected.

Conclusion/Significance

Both BP artificial neural network and logistic regression model established in a small scale suggested that individual behavior and socioeconomic status are the most important risk factors in the transmission of schistosomiasis japonica. It was reviewed that the young population (≤15) in higher-risk areas was the main target to be intervened for the disease transmission control.     

Deep Convolution Neural Network Based System for Early Diagnosis of Alzheimer's Disease

ObjectivesAlzheimer's Disease (AD) is the most general type of dementia. In all leading countries, it is one of the primary reasons of death in senior citizens. Currently, it is diagnosed by calculating the MSME score and by the manual study of MRI Scan. Also, different machine learning methods are utilized for automatic diagnosis but existing has some limitations in terms of accuracy. So, main objective of this paper to include a preprocessing method before CNN model to increase the accuracy of classification.Materials and methodIn this paper, we present a deep learning-based approach for detection of Alzheimer's Disease from ADNI database of Alzheimer's disease patients, the dataset contains fMRI and PET images of Alzheimer's patients along with normal person's image. We have applied 3D to 2D conversion and resizing of images before applying VGG-16 architecture of Convolution neural network for feature extraction. Finally, for classification SVM, Linear Discriminate, K means clustering, and Decision tree classifiers are used.ResultsThe experimental result shows that the average accuracy of 99.95% is achieved for the classification of the fMRI dataset, while the average accuracy of 73.46% is achieved with the PET dataset. On comparing results on the basis of accuracy, specificity, sensitivity and on some other parameters we found that these results are better than existing methods.Conclusionsthis paper, suggested a unique way to increase the performance of CNN models by applying some preprocessing on image dataset before sending to CNN architecture for feature extraction. We applied this method on ADNI database and on comparing the accuracies with other similar approaches it shows better results.     

A Multi-Atlas Based Method for Automated Anatomical Rat Brain MRI Segmentation and Extraction of PET Activity

Introduction

Preclinical in vivo imaging requires precise and reproducible delineation of brain structures. Manual segmentation is time consuming and operator dependent. Automated segmentation as usually performed via single atlas registration fails to account for anatomo-physiological variability. We present, evaluate, and make available a multi-atlas approach for automatically segmenting rat brain MRI and extracting PET activies.

Methods

High-resolution 7T 2DT2 MR images of 12 Sprague-Dawley rat brains were manually segmented into 27-VOI label volumes using detailed protocols. Automated methods were developed with 7/12 atlas datasets, i.e. the MRIs and their associated label volumes. MRIs were registered to a common space, where an MRI template and a maximum probability atlas were created. Three automated methods were tested: 1/registering individual MRIs to the template, and using a single atlas (SA), 2/using the maximum probability atlas (MP), and 3/registering the MRIs from the multi-atlas dataset to an individual MRI, propagating the label volumes and fusing them in individual MRI space (propagation & fusion, PF). Evaluation was performed on the five remaining rats which additionally underwent [¹⁸F]FDG PET. Automated and manual segmentations were compared for morphometric performance (assessed by comparing volume bias and Dice overlap index) and functional performance (evaluated by comparing extracted PET measures).

Results

Only the SA method showed volume bias. Dice indices were significantly different between methods (PF>MP>SA). PET regional measures were more accurate with multi-atlas methods than with SA method.

Conclusions

Multi-atlas methods outperform SA for automated anatomical brain segmentation and PET measure’s extraction. They perform comparably to manual segmentation for FDG-PET quantification. Multi-atlas methods are suitable for rapid reproducible VOI analyses.     

Probabilistic Interaction Network of Evidence Algorithm and its Application to Complete Labeling of Peak Lists from Protein NMR Spectroscopy

The process of assigning a finite set of tags or labels to a collection of observations, subject to side conditions, is notable for its computational complexity. This labeling paradigm is of theoretical and practical relevance to a wide range of biological applications, including the analysis of data from DNA microarrays, metabolomics experiments, and biomolecular nuclear magnetic resonance (NMR) spectroscopy. We present a novel algorithm, called Probabilistic Interaction Network of Evidence (PINE), that achieves robust, unsupervised probabilistic labeling of data. The computational core of PINE uses estimates of evidence derived from empirical distributions of previously observed data, along with consistency measures, to drive a fictitious system M with Hamiltonian H to a quasi-stationary state that produces probabilistic label assignments for relevant subsets of the data. We demonstrate the successful application of PINE to a key task in protein NMR spectroscopy: that of converting peak lists extracted from various NMR experiments into assignments associated with probabilities for their correctness. This application, called PINE-NMR, is available from a freely accessible computer server (http://pine.nmrfam.wisc.edu). The PINE-NMR server accepts as input the sequence of the protein plus user-specified combinations of data corresponding to an extensive list of NMR experiments; it provides as output a probabilistic assignment of NMR signals (chemical shifts) to sequence-specific backbone and aliphatic side chain atoms plus a probabilistic determination of the protein secondary structure. PINE-NMR can accommodate prior information about assignments or stable isotope labeling schemes. As part of the analysis, PINE-NMR identifies, verifies, and rectifies problems related to chemical shift referencing or erroneous input data. PINE-NMR achieves robust and consistent results that have been shown to be effective in subsequent steps of NMR structure determination.     

A Spike-Timing Pattern Based Neural Network Model for the Study of Memory Dynamics

It is well accepted that the brain''s computation relies on spatiotemporal activity of neural networks. In particular, there is growing evidence of the importance of continuously and precisely timed spiking activity. Therefore, it is important to characterize memory states in terms of spike-timing patterns that give both reliable memory of firing activities and precise memory of firing timings. The relationship between memory states and spike-timing patterns has been studied empirically with large-scale recording of neuron population in recent years. Here, by using a recurrent neural network model with dynamics at two time scales, we construct a dynamical memory network model which embeds both fast neural and synaptic variation and slow learning dynamics. A state vector is proposed to describe memory states in terms of spike-timing patterns of neural population, and a distance measure of state vector is defined to study several important phenomena of memory dynamics: partial memory recall, learning efficiency, learning with correlated stimuli. We show that the distance measure can capture the timing difference of memory states. In addition, we examine the influence of network topology on learning ability, and show that local connections can increase the network''s ability to embed more memory states. Together theses results suggest that the proposed system based on spike-timing patterns gives a productive model for the study of detailed learning and memory dynamics.     

A Rapid Procedure for the Large Scale Purification of Elastase and Cathepsin G from Human Sputum

A procedure is described which permits the rapid isolation of large amounts of elastase and cathepsin G from purulent sputum. This procedure involves: (1) digestion of sputum with DNase, (2) extraction of the insoluble residue that remains with 1 M NaCl, pH 8, (3) affinity chromatography on Sepharose-bound Trasylol, and (4) separation of the two enzymes by chromatography on CM-Sephadex. Starting with 500 g of sputum it was possible to isolate 175 mg of each of these two enzymes within 7 to 10 days. Active site titration indicated both enzymes to be at least 97% pure. Disc gel electrophoresis in the presence and absence of SDS and amino acid sequence of the N-terminal region support the conclusion that the elastase and cathepsin G isolated from sputum a re identical to the same enzymes isolated directly from the leukocytes of human blood.     

Virtual Power Based Algorithm for Decoupling Large Motions from Infinitesimal Strains: Application to Shoulder Joint Biomechanics

New trends of numerical models of human joints require more and more computation of both large amplitude joint motions and fine bone stress distribution. Together, these problems are difficult to solve and very CPU time consuming. The goal of this study is to develop a new method to diminish the calculation time for this kind of problems which include calculation of large amplitude motions and infinitesimal strains. Based on the Principle of Virtual Power, the present method decouples the problem into two parts. First, rigid body motion is calculated. The bone micro-deformations are then calculated in a second part by using the results of rigid body motions as boundary conditions. A finite element model of the shoulder was used to test this decoupling technique. The model was designed to determine the influence of humeral head shape on stress distribution in the scapula for different physiological motions of the joint. Two versions of the model were developed: a first version completely deformable and a second version based on the developed decoupling method. It was shown that biomechanical variables, as mean pressure and von Mises stress, calculated with the two versions were sensibly the same. On the other hand, CPU time needed for calculating with the new decoupled technique was more than 6 times less than with the completely deformable model.     

Application of Artificial Neural Network for predicting biomass growth during domestic wastewater treatment through a biological process

The biological treatment process is responsible for removing organic and inorganic matter in wastewater. This process relies heavily on microorganisms to successfully remove organic and inorganic matter. The aim of the study was to model biomass growth in the biological treatment process. Multilayer perceptron (MLP) Artificial Neural Network (ANN) algorithm was used to model biomass growth. Three metrics: coefficient of determination (R²), root mean squared error (RMSE), and mean squared error (MSE) were used to evaluate the performance of the model. Sensitivity analysis was applied to confirm variables that have a strong influence on biomass growth. The results of the study showed that MLP ANN algorithm was able to model biomass growth successfully. R² values were 0.844, 0.853, and 0.823 during training, validation, and testing phases, respectively. RMSE values were 0.7476, 1.1641, and 0.7798 during training, validation, and testing phases respectively. MSE values were 0.5589, 1.3551, and 0.6081 during training, validation, and testing phases, respectively. Sensitivity analysis results showed that temperature (47.2%) and dissolved oxygen (DO) concentration (40.2%) were the biggest drivers of biomass growth. Aeration period (4.3%), chemical oxygen demand (COD) concentration (3.2%), and oxygen uptake rate (OUR) (5.1%) contributed minimally. The biomass growth model can be applied at different wastewater treatment plants by different plant managers/operators in order to achieve optimum biomass growth. The optimum biomass growth will improve the removal of organic and inorganic matter in the biological treatment process.     

Modification of the Starch Block Electrophoresis Method for the Preparation of Immunoglobulin A from Multiple Myeloma Sera on a Large Scale

Separation of immunoglobulin A from a large sample of human serum from patients with multiple myeloma was done by utilizing a large zone electrophoresis apparatus.     

A Multi-Branch Convolutional Neural Network for Screening and Staging of Diabetic Retinopathy Based on Wide-Field Optical Coherence Tomography Angiography

BackgroundDiabetic retinopathy (DR) is one of the major causes of blindness in adults suffering from diabetes. With the development of wide-field optical coherence tomography angiography (WF-OCTA), it is to become a gold standard for diagnosing DR. The demand for automated DR diagnosis system based on OCTA images have been fostered due to large diabetic population and pervasiveness of retinopathy cases.Materials and methodsIn this study, 288 diabetic patients and 97 healthy people were imaged by the swept-source optical coherence tomography (SS-OCT) with 12 mm × 12 mm single scan centered on the fovea. A multi-branch convolutional neural network (CNN) was proposed to classify WF-OCTA images into four grades: no DR, mild non-proliferative diabetic retinopathy (NPDR), moderate to severe NPDR, and proliferative diabetic retinopathy (PDR).ResultsThe proposed model achieved a classification accuracy of 96.11%, sensitivity of 98.08% and specificity of 89.43% in detecting DR. The accuracy of the model for DR staging is 90.56%, which is higher than that of other mainstream convolution neural network models.ConclusionThis technology enables early diagnosis and objective tracking of disease progression, which may be useful for optimal treatment to reduce vision loss.     

Application of Physicochemical Properties and Process Parameters in the Development of a Neural Network Model for Prediction of Tablet Characteristics

The importance of in silico modeling in the pharmaceutical industry is continuously increasing. The aim of the present study was the development of a neural network model for prediction of the postcompressional properties of scored tablets based on the application of existing data sets from our previous studies. Some important process parameters and physicochemical characteristics of the powder mixtures were used as training factors to achieve the best applicability in a wide range of possible compositions. The results demonstrated that, after some pre-processing of the factors, an appropriate prediction performance could be achieved. However, because of the poor extrapolation capacity, broadening of the training data range appears necessary.     

Application of Network Scale Up Method in the Estimation of Population Size for Men Who Have Sex with Men in Shanghai,China

Background

Men who have sex with men (MSM) are at high risk of HIV infection. For developing proper interventions, it is important to know the size of MSM population. However, size estimation of MSM populations is still a significant public health challenge due to high cost, hard to reach and stigma associated with the population.

Objectives

We aimed to estimate the social network size (c value) in general population and the size of MSM population in Shanghai, China by using the net work scale-up method.

Methods

A multistage random sampling was used to recruit participants aged from 18 to 60 years who had lived in Shanghai for at least 6 months. The “known population method” with adjustment of backward estimation and regression model was applied to estimate the c value. And the MSM population size was further estimated using an adjusted c value taking into account for the transmission effect through social respect level towards MSM.

Results

A total of 4017 participants were contacted for an interview, and 3907 participants met the inclusion criterion. The social network size (c value) of participants was 236 after adjustment. The estimated size of MSM was 36354 (95% CI: 28489–44219) for the male Shanghaies aged 18 to 60 years, and the proportion of MSM among the total male population aged 18 to 60 years in Shanghai was 0.28%.

Conclusions

We employed the network scale-up method and used a wide range of data sources to estimate the size of MSM population in Shanghai, which is useful for HIV prevention and intervention among the target population.     

A Method for DNA Extraction from the Desert Cyanobacterium Chroococcidiopsis and Its Application to Identification of ftsZ

A method was developed for extraction of DNA from Chroococcidiopsis that overcomes obstacles posed by bacterial contamination and the presence of a thick envelope surrounding the cyanobacterial cells. The method is based on the resistance of Chroococcidiopsis to lysozyme and consists of a lysozyme treatment followed by osmotic shock that reduces the bacterial contamination by 3 orders of magnitude. Then DNase treatment is performed to eliminate DNA from the bacterial lysate. Lysis of Chroococcidiopsis cells is achieved by grinding with glass beads in the presence of hot phenol. Extracted DNA is further purified by cesium-chloride density gradient ultracentrifugation. This method permitted the first molecular approach to the study of Chroococcidiopsis, and a 570-bp fragment of the gene ftsZ was cloned and sequenced.     

A Concise Review of Extraction and Characterization of Chondroitin Sulphate from Fish and Fish Wastes for Pharmacological Application

Chondroitin sulphate (CS) is one of the most predominant glycosaminoglycans (GAGs) available in the extracellular matrix of tissues. It has many health benefits, including relief from osteoarthritis, antiviral properties, tissue engineering applications, and use in skin care, which have increased its commercial demand in recent years. The quest for CS sources exponentially increased due to several shortcomings of porcine, bovine, and other animal sources. Fish and fish wastes (i.e., fins, scales, skeleton, bone, and cartilage) are suitable sources of CS as they are low cost, easy to handle, and readily available. However, the lack of a standard isolation and characterization technique makes CS production challenging, particularly concerning the yield of pure GAGs. Many studies imply that enzyme-based extraction is more effective than chemical extraction. Critical evaluation of the existing extraction, isolation, and characterization techniques is crucial for establishing an optimized protocol of CS production from fish sources. The current techniques depend on tissue hydrolysis, protein removal, and purification. Therefore, this study critically evaluated and discussed the extraction, isolation, and characterization methods of CS from fish or fish wastes. Biosynthesis and pharmacological applications of CS were also critically reviewed and discussed. Our assessment suggests that CS could be a potential drug candidate; however, clinical studies should be conducted to warrant its effectiveness.     

New Method for Extraction of Streptomycete Spores from Soil and Application to the Study of Lysogeny in Sterile Amended and Nonsterile Soil

A new method for the isolation and enumeration of streptomycete spores from soil was developed. This method makes use of a cation-exchange resin to disperse soil particles. It allowed the detection of 10 spores in 100 g of sterile soil, while ca. 10³ could be accurately enumerated in 100 g. This method was applied to studying the fate of a marked actinophage in soil. In sterile amended and nonsterile soil, relatively high numbers of actinophages were only found during the first few days of the experiment when the host streptomycete was in the mycelial form. Later, after sporulation, lysogens could be detected in sterile amended soil and could still be found 60 days after inoculation. Although no lysogens were found in nonsterile soil, the introduced phage could still be detected in the free state after 60 days, albeit at a low titer.     

Rapid Method for Isolation of Large Quantities of Outer Membrane from Escherichia coli K-12 and Its Application to the Study of Envelope Mutants

A rapid method for the isolation of large quantities of bacterial outer membrane is described. This cell envelope component was removed from plasmolyzed cells of Escherichia coli K-12 by lysozyme-ethylenediaminetetra-acetic acid treatment, aggregated by lowering the pH to 5.0, and recovered by centrifugation. Aggregates of membrane fragments were clearly identified in an electron microscope. A criterion of homogeneity of the preparation was obtained by isopycnic sucrose gradient centrifugation. A single band appeared at a density of 1.24 g/cc. The cytoplasmic membrane marker, succinate dehydrogenase activity, was 40 times lower in the outer membrane preparation than in complete cell envelope preparations. A rich activity was, however, found for the outer membrane marker, phospholipase A. The compositions of outer membranes from a transductant pair were compared. One transductant was a chain-forming, antibiotic-supersensitive envA strain, whereas the other contained the envA(+) allele. The envA strain showed a slightly modified protein pattern and a lower relative content of phosphatidylglycerol.