Predicting Cardiotoxicity of Molecules Using Attention-Based Graph Neural Networks

In drug discovery, the search for new and effective medications is often hindered by concerns about toxicity. Numerous promising molecules fail to pass the later phases of drug development due to strict toxicity assessments. This challenge significantly increases the cost, time, and human effort needed to discover new therapeutic molecules. Additionally, a considerable number of drugs already on the market have been withdrawn or re-evaluated because of their unwanted side effects. Among the various types of toxicity, drug-induced heart damage is a severe adverse effect commonly associated with several medications, especially those used in cancer treatments. Although a number of computational approaches have been proposed to identify the cardiotoxicity of molecules, the performance and interpretability of the existing approaches are limited. In our study, we proposed a more effective computational framework to predict the cardiotoxicity of molecules using an attention-based graph neural network. Experimental results indicated that the proposed framework outperformed the other methods. The stability of the model was also confirmed by our experiments. To assist researchers in evaluating the cardiotoxicity of molecules, we have developed an easy-to-use online web server that incorporates our model.

Herpes zoster vaccine safety in the Aotearoa New Zealand population: a self-controlled case series study

In Aotearoa New Zealand, zoster vaccine live is used for the prevention of zoster and associated complications in adults. This study assessed the risk of pre-specified serious adverse events following zoster vaccine live immunisation among adults in routine clinical practice. We conducted a self-controlled case series study using routinely collected national data. We compared the incidence of serious adverse events during the at-risk period with the control period. Rate ratios were estimated using Conditional Poisson regression models. Falsification outcomes analyses were used to evaluate biases in our study population. From April 2018 to July 2021, 278,375 received the vaccine. The rate ratio of serious adverse events following immunisation was 0·43 (95% confidence interval [CI]: 0·37-0·50). There was no significant increase in the risk of cerebrovascular accidents, acute myocardial infarction, acute pericarditis, acute myocarditis, and Ramsay-Hunt Syndrome. The herpes zoster vaccine is safe in adults in Aotearoa New Zealand.

i4mC-GRU: Identifying DNA N4-Methylcytosine sites in mouse genomes using bidirectional gated recurrent unit and sequence-embedded features

N⁴-methylcytosine (4mC) is one of the most common DNA methylation modifications found in both prokaryotic and eukaryotic genomes. Since the 4mC has various essential biological roles, determining its location helps reveal unexplored physiological and pathological pathways. In this study, we propose an effective computational method called i4mC-GRU using a gated recurrent unit and duplet sequence-embedded features to predict potential 4mC sites in mouse (Mus musculus) genomes. To fairly assess the performance of the model, we compared our method with several state-of-the-art methods using two different benchmark datasets. Our results showed that i4mC-GRU achieved area under the receiver operating characteristic curve values of 0.97 and 0.89 and area under the precision-recall curve values of 0.98 and 0.90 on the first and second benchmark datasets, respectively. Briefly, our method outperformed existing methods in predicting 4mC sites in mouse genomes. Also, we deployed i4mC-GRU as an online web server, supporting users in genomics studies.

Predicting high health-cost users among people with cardiovascular disease using machine learning and nationwide linked social administrative datasets

OBJECTIVES

To optimise planning of public health services, the impact of high-cost users needs to be considered. However, most of the existing statistical models for costs do not include many clinical and social variables from administrative data that are associated with elevated health care resource use, and are increasingly available. This study aimed to use machine learning approaches and big data to predict high-cost users among people with cardiovascular disease (CVD).

METHODS

We used nationally representative linked datasets in New Zealand to predict CVD prevalent cases with the most expensive cost belonging to the top quintiles by cost. We compared the performance of four popular machine learning models (L1-regularised logistic regression, classification trees, k-nearest neighbourhood (KNN) and random forest) with the traditional regression models.

RESULTS

The machine learning models had far better accuracy in predicting high health-cost users compared with the logistic models. The harmony score F1 (combining sensitivity and positive predictive value) of the machine learning models ranged from 30.6% to 41.2% (compared with 8.6-9.1% for the logistic models). Previous health costs, income, age, chronic health conditions, deprivation, and receiving a social security benefit were among the most important predictors of the CVD high-cost users.

CONCLUSIONS

This study provides additional evidence that machine learning can be used as a tool together with big data in health economics for identification of new risk factors and prediction of high-cost users with CVD. As such, machine learning may potentially assist with health services planning and preventive measures to improve population health while potentially saving healthcare costs.

Herpes zoster vaccine effectiveness against herpes zoster and postherpetic neuralgia in New Zealand: a retrospective cohort study

BACKGROUND

Herpes zoster (HZ) and associated complications cause significant burden to older people. A HZ vaccination programme was introduced in Aotearoa New Zealand in April 2018 with a single dose vaccine for those aged 65 years and a four-year catch up for 66-80 year-olds. This study aimed to assess the 'real-world' effectiveness of the zoster vaccine live (ZVL) against HZ and postherpetic neuralgia (PHN).

METHODS

We conducted a nationwide retrospective matched cohort study from 1 April 2018 to 1 April 2021 using a linked de-identified patient level Ministry of Health data platform. A Cox proportional hazards model was used to estimate ZVL vaccine effectiveness (VE) against HZ and PHN adjusting for covariates. Multiple outcomes were assessed in the primary (hospitalised HZ and PHN - primary diagnosis) and secondary (hospitalised HZ and PHN: primary and secondary diagnosis, community HZ) analyses. A sub-group analysis was carried out in, adults ≥ 65 years old, immunocompromised adults, Māori, and Pacific populations.

FINDINGS

A total of 824,142 (274,272 vaccinated with ZVL matched with 549,870 unvaccinated) New Zealand residents were included in the study. The matched population was 93.4% immunocompetent, 52.2% female, 80.2% European (level 1 ethnic codes), and 64.5% were 65-74 years old (mean age = 71.1±5.0). Vaccinated versus unvaccinated incidence of hospitalised HZ was 0.16 vs. 0.31/1,000 person-years and 0.03 vs. 0.08/1000 person-years for PHN. In the primary analysis, the adjusted overall VE against hospitalised HZ and hospitalised PHN was 57.8% (95% CI: 41.1-69.8) and 73.7% (95% CI:14.0-92.0) respectively. In adults ≥ 65 years old, the VE against hospitalised HZ was 54.4% (95% CI: 36.0-67.5) and VE against hospitalised PHN was 75·5% (95% CI: 19.9-92.5). In the secondary analysis, the VE against community HZ was 30.0% (95% CI: 25.6-34.5). The ZVL VE against hospitalised HZ for immunocompromised adults was 51.1% (95% CI: 23.1-69.5), and PHN hospitalisation was 67.6% (95% CI: 9.3-88.4). The VE against HZ hospitalisation for Māori was 45.2% (95% CI: -23.2-75.6) and for Pacific Peoples was 52.2% (95% CI: -40.6 -83·7).

INTERPRETATION

ZVL was associated with a reduction in risk of hospitalisation from HZ and PHN in the New Zealand population.

FUNDING

Wellington Doctoral Scholarship awarded to JFM.

iNSP-GCAAP: Identifying nonclassical secreted proteins using global composition of amino acid properties

Nonclassical secreted proteins (NSPs) refer to a group of proteins released into the extracellular environment under the facilitation of different biological transporting pathways apart from the Sec/Tat system. As experimental determination of NSPs is often costly and requires skilled handling techniques, computational approaches are necessary. In this study, we introduce iNSP-GCAAP, a computational prediction framework, to identify NSPs. We propose using global composition of a customized set of amino acid properties to encode sequence data and use the random forest (RF) algorithm for classification. We used the training dataset introduced by Zhang et al. (Bioinformatics, 36(3), 704-712, 2020) to develop our model and test it with the independent test set in the same study. The area under the receiver operating characteristic curve on that test set was 0.9256, which outperformed other state-of-the-art methods using the same datasets. Our framework is also deployed as a user-friendly web-based application to support the research community to predict NSPs.

eMIC-AntiKP: Estimating minimum inhibitory concentrations of antibiotics towards Klebsiella pneumoniae using deep learning

Nowadays, antibiotic resistance has become one of the most concerning problems that directly affects the recovery process of patients. For years, numerous efforts have been made to efficiently use antimicrobial drugs with appropriate doses not only to exterminate microbes but also stringently constrain any chances for bacterial evolution. However, choosing proper antibiotics is not a straightforward and time-effective process because well-defined drugs can only be given to patients after determining microbic taxonomy and evaluating minimum inhibitory concentrations (MICs). Besides conventional methods, numerous computer-aided frameworks have been recently developed using computational advances and public data sources of clinical antimicrobial resistance. In this study, we introduce eMIC-AntiKP, a computational framework specifically designed to predict the MIC values of 20 antibiotics towards Klebsiella pneumoniae. Our prediction models were constructed using convolutional neural networks and k-mer counting-based features. The model for cefepime has the most limited performance with a test 1-tier accuracy of 0.49, while the model for ampicillin has the highest performance with a test 1-tier accuracy of 1.00. Most models have satisfactory performance, with test accuracies ranging from about 0.70-0.90. The significance of eMIC-AntiKP is the effective utilization of computing resources to make it a compact and portable tool for most moderately configured computers. We provide users with two options, including an online web server for basic analysis and an offline package for deeper analysis and technical modification.

A lightweight classification of adaptor proteins using transformer networks

BACKGROUND

Adaptor proteins play a key role in intercellular signal transduction, and dysfunctional adaptor proteins result in diseases. Understanding its structure is the first step to tackling the associated conditions, spurring ongoing interest in research into adaptor proteins with bioinformatics and computational biology. Our study aims to introduce a small, new, and superior model for protein classification, pushing the boundaries with new machine learning algorithms.

RESULTS

We propose a novel transformer based model which includes convolutional block and fully connected layer. We input protein sequences from a database, extract PSSM features, then process it via our deep learning model. The proposed model is efficient and highly compact, achieving state-of-the-art performance in terms of area under the receiver operating characteristic curve, Matthew's Correlation Coefficient and Receiver Operating Characteristics curve. Despite merely 20 hidden nodes translating to approximately 1% of the complexity of previous best known methods, the proposed model is still superior in results and computational efficiency.

CONCLUSIONS

The proposed model is the first transformer model used for recognizing adaptor protein, and outperforms all existing methods, having PSSM profiles as inputs that comprises convolutional blocks, transformer and fully connected layers for the use of classifying adaptor proteins.

iPromoter-Seqvec: identifying promoters using bidirectional long short-term memory and sequence-embedded features

Background

Promoters, non-coding DNA sequences located at upstream regions of the transcription start site of genes/gene clusters, are essential regulatory elements for the initiation and regulation of transcriptional processes. Furthermore, identifying promoters in DNA sequences and genomes significantly contributes to discovering entire structures of genes of interest. Therefore, exploration of promoter regions is one of the most imperative topics in molecular genetics and biology. Besides experimental techniques, computational methods have been developed to predict promoters. In this study, we propose iPromoter-Seqvec – an efficient computational model to predict TATA and non-TATA promoters in human and mouse genomes using bidirectional long short-term memory neural networks in combination with sequence-embedded features extracted from input sequences. The promoter and non-promoter sequences were retrieved from the Eukaryotic Promoter database and then were refined to create four benchmark datasets.

Results

The area under the receiver operating characteristic curve (AUCROC) and the area under the precision-recall curve (AUCPR) were used as two key metrics to evaluate model performance. Results on independent test sets showed that iPromoter-Seqvec outperformed other state-of-the-art methods with AUCROC values ranging from 0.85 to 0.99 and AUCPR values ranging from 0.86 to 0.99. Models predicting TATA promoters in both species had slightly higher predictive power compared to those predicting non-TATA promoters. With a novel idea of constructing artificial non-promoter sequences based on promoter sequences, our models were able to learn highly specific characteristics discriminating promoters from non-promoters to improve predictive efficiency.

Conclusions

iPromoter-Seqvec is a stable and robust model for predicting both TATA and non-TATA promoters in human and mouse genomes. Our proposed method was also deployed as an online web server with a user-friendly interface to support research communities. Links to our source codes and web server are available at https://github.com/mldlproject/2022-iPromoter-Seqvec.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12864-022-08829-6.

Identifying Transcription Factors That Prefer Binding to Methylated DNA Using Reduced G-Gap Dipeptide Composition

Transcription factors (TFs) play an important role in gene expression and regulation of 3D genome conformation. TFs have ability to bind to specific DNA fragments called enhancers and promoters. Some TFs bind to promoter DNA fragments which are near the transcription initiation site and form complexes that allow polymerase enzymes to bind to initiate transcription. Previous studies showed that methylated DNAs had ability to inhibit and prevent TFs from binding to DNA fragments. However, recent studies have found that there were TFs that could bind to methylated DNA fragments. The identification of these TFs is an important steppingstone to a better understanding of cellular gene expression mechanisms. However, as experimental methods are often time-consuming and labor-intensive, developing computational methods is essential. In this study, we propose two machine learning methods for two problems: (1) identifying TFs and (2) identifying TFs that prefer binding to methylated DNA targets (TFPMs). For the TF identification problem, the proposed method uses the position-specific scoring matrix for data representation and a deep convolutional neural network for modeling. This method achieved 90.56% sensitivity, 83.96% specificity, and an area under the receiver operating characteristic curve (AUC) of 0.9596 on an independent test set. For the TFPM identification problem, we propose to use the reduced g-gap dipeptide composition for data representation and the support vector machine algorithm for modeling. This method achieved 82.61% sensitivity, 64.86% specificity, and an AUC of 0.8486 on another independent test set. These results are higher than those of other studies on the same problems.

Post-licensure zoster vaccine effectiveness against herpes zoster and postherpetic neuralgia in older adults: a systematic review and meta-analysis

BACKGROUND

Given the substantial impact of herpes zoster on health and quality of life, and its considerable economic burden, prevention through vaccination is a priority. We aimed to evaluate the effectiveness of the herpes zoster vaccines (recombinant zoster vaccine [RZV] and zoster vaccine live [ZVL]) against incident herpes zoster and postherpetic neuralgia in older adults.

METHODS

We did a systematic review and meta-analysis of studies assessing the effectiveness of herpes zoster vaccines in adults aged 50 years or older, compared with no vaccination or another vaccine. We searched published literature on MEDLINE, Embase, Cochrane Library, Cumulative Index to Nursing and Allied Health Literature, ProQuest Central, and Dimensions, as well as unpublished studies, grey literature, and the reference lists of included studies. Observational studies published in any language between May 25, 2006, and Dec 31, 2020, were included. Eligible studies were appraised for methodological quality using standardised critical appraisal instruments from the Joanna Briggs Institute, and data were extracted from selected studies using a standardised tool. Random-effects meta-analysis models were used to estimate pooled vaccine effectiveness for outcomes of interest (herpes zoster, herpes zoster ophthalmicus, and postherpetic neuralgia) among clinically and methodologically comparable studies, with a fixed-effects model also used for herpes zoster ophthalmicus. Vaccine effectiveness was also assessed in people with comorbidities. As a post-hoc analysis, a forward citation search was done on Jan 31, 2021. This study is registered on PROSPERO, CRD42021232383.

FINDINGS

Our search identified 1240 studies, of which 1162 were excluded based on title and abstract screening. A further 56 articles were excluded on reading the full text. 22 studies (21 cohort studies and one case-control study, involving 9 536 086 participants and 3·35 million person-years in the USA, UK, Canada, and Sweden) were included in the quantitative analysis. Of these, 13 articles were included in the meta-analysis. The overall quality of evidence was very low for all outcomes. The pooled vaccine effectiveness for ZVL against herpes zoster in adults was 45·9% (95% CI 42·2-49·4; seven studies). The vaccine effectiveness for ZVL against postherpetic neuralgia was 59·7% (58·4-89·7; three studies) and against herpes zoster ophthalmicus (in a fixed-effects model) was 30·0% (20·5-38·4; two studies). ZVL was effective in preventing herpes zoster in people with comorbidities, including diabetes (vaccine effectiveness 49·8%, 45·1-54·1; three studies), chronic kidney disease (54·3%, 49·0-59·1; four studies), liver disease (52·9%, 41·6-62·1; two studies), heart disease (52·3%, 45·0-58·7; two studies), and lung disease (49·0%, 32·2-66·2; two studies). In a post-hoc analysis of two studies from the USA published after 2020, the pooled vaccine effectiveness for RZV against herpes zoster in adults was 79·2% (57·6-89·7). Substantial heterogeneity (I²≥75%) was observed in 50% of the meta-analyses.

INTERPRETATION

ZVL and RZV are effective in preventing herpes zoster in routine clinical practice. ZVL also reduces the risk of postherpetic neuralgia. Selection bias and confounding by unmeasured variables are inherent challenges of observational studies based on large health-care databases. Nevertheless, these findings will reassure policy makers, health practitioners, and the public that the vaccinations currently available for herpes zoster vaccination programmes are effective at preventing herpes zoster and related complications.

FUNDING

None.

Application of machine learning in screening for congenital heart diseases using fetal echocardiography

There is a growing body of literature supporting the utilization of machine learning (ML) to improve diagnosis and prognosis tools of cardiovascular disease. The current study was to investigate the impact that the ML framework may have on the sensitivity of predicting the presence or absence of congenital heart disease (CHD) using fetal echocardiography. A comprehensive fetal echocardiogram including 2D cardiac chamber quantification, valvar assessments, assessment of great vessel morphology, and Doppler-derived blood flow interrogation was recorded. The postnatal echocardiogram was used to ascertain the diagnosis of CHD. A random forest (RF) algorithm with a nested tenfold cross-validation was used to train models for assessing the presence of CHD. The study population was derived from a database of 3910 singleton fetuses with maternal age of 28.8 ± 5.2 years and gestational age at the time of fetal echocardiography of 22.0 weeks (IQR 21-24). The proportion of CHD was 14.1% for the studied cohort confirmed by post-natal echocardiograms. Our proposed RF-based framework provided a sensitivity of 0.85, a specificity of 0.88, a positive predictive value of 0.55 and a negative predictive value of 0.97 to detect the CHD with the mean of mean ROC curves of 0.94 and the mean of mean PR curves of 0.84. Additionally, six first features, including cardiac axis, peak velocity of blood flow across the pulmonic valve, cardiothoracic ratio, pulmonary valvar annulus diameter, right ventricular end-diastolic diameter, and aortic valvar annulus diameter, are essential features that play crucial roles in adding more predictive values to the model in detecting patients with CHD. ML using RF can provide increased sensitivity in prenatal CHD screening with very good performance. The incorporation of ML algorithms into fetal echocardiography may further standardize the assessment for CHD.

iANP-EC: Identifying Anticancer Natural Products Using Ensemble Learning Incorporated with Evolutionary Computation

Cancer is one of the most deadly diseases that annually kills millions of people worldwide. The investigation on anticancer medicines has never ceased to seek better and more adaptive agents with fewer side effects. Besides chemically synthetic anticancer compounds, natural products are scientifically proved as a highly potential alternative source for anticancer drug discovery. Along with experimental approaches being used to find anticancer drug candidates, computational approaches have been developed to virtually screen for potential anticancer compounds. In this study, we construct an ensemble computational framework, called iANP-EC, using machine learning approaches incorporated with evolutionary computation. Four learning algorithms (k-NN, SVM, RF, and XGB) and four molecular representation schemes are used to build a set of classifiers, among which the top-four best-performing classifiers are selected to form an ensemble classifier. Particle swarm optimization (PSO) is used to optimise the weights used to combined the four top classifiers. The models are developed by a set of curated 997 compounds which are collected from the NPACT and CancerHSP databases. The results show that iANP-EC is a stable, robust, and effective framework that achieves an AUC-ROC value of 0.9193 and an AUC-PR value of 0.8366. The comparative analysis of molecular substructures between natural anticarcinogens and nonanticarcinogens partially unveils several key substructures that drive anticancerous activities. We also deploy the proposed ensemble model as an online web server with a user-friendly interface to support the research community in identifying natural products with anticancer activities.

Prediction of FMN Binding Sites in Electron Transport Chains Based on 2-D CNN and PSSM Profiles

Flavin mono-nucleotides (FMNs) are cofactors that hold responsibility for carrying and transferring electrons in the electron transport chain stage of cellular respiration. Without being facilitated by FMNs, energy production is stagnant due to the interruption in most of the cellular processes. Investigation on FMN's functions, therefore, can gain holistic understanding about human diseases and molecular information on drug targets. We proposed a deep learning model using a two-dimensional convolutional neural network and position specific scoring matrices that could identify FMN interacting residues with the sensitivity of 83.7 percent, specificity of 99.2 percent, accuracy of 98.2 percent, and Matthews correlation coefficients of 0.85 for an independent dataset containing 141 FMN binding sites and 1,920 non-FMN binding sites. The proposed method outperformed other previous studies using similar evaluation metrics. Our positive outcome can also promote the utilization of deep learning in dealing with various problems in bioinformatics and computational biology.

iCYP-MFE: Identifying Human Cytochrome P450 Inhibitors Using Multitask Learning and Molecular Fingerprint-Embedded Encoding

The human cytochrome P450 (CYP) superfamily holds responsibilities for the metabolism of both endogenous and exogenous compounds such as drugs, cellular metabolites, and toxins. The inhibition exerted on the CYP enzymes is closely associated with adverse drug reactions encompassing metabolic failures and induced side effects. In modern drug discovery, identification of potential CYP inhibitors is, therefore, highly essential. Alongside experimental approaches, numerous computational models have been proposed to address this biochemical issue. In this study, we introduce iCYP-MFE, a computational framework for virtual screening on CYP inhibitors toward 1A2, 2C9, 2C19, 2D6, and 3A4 isoforms. iCYP-MFE contains a set of five robust, stable, and effective prediction models developed using multitask learning incorporated with molecular fingerprint-embedded features. The results show that multitask learning can remarkably leverage useful information from related tasks to promote global performance. Comparative analysis indicates that iCYP-MFE achieves three predominant tasks, one equivalent task, and one less effective task compared to state-of-the-art methods. The area under the receiver operating characteristic curve (AUC-ROC) and the area under the precision-recall curve (AUC-PR) were two decisive metrics used for model evaluation. The prediction task for CYP2D6-inhibition achieves the highest AUC-ROC value of 0.93 while the prediction task for CYP1A2-inhibition obtains the highest AUC-PR value of 0.92. The substructural analysis preliminarily explains the nature of the CYP-inhibitory activity of compounds. An online web server for iCYP-MFE with a user-friendly interface was also deployed to support scientific communities in identifying CYP inhibitors.

COVID-19 vaccine strategies for Aotearoa New Zealand: a mathematical modelling study

Background: COVID-19 elimination measures, including border closures have been applied in New Zealand. We have modelled the potential effect of vaccination programmes for opening borders. Methods: We used a deterministic age-stratified Susceptible, Exposed, Infectious, Recovered (SEIR) model. We minimised spread by varying the age-stratified vaccine allocation to find the minimum herd immunity requirements (the effective reproduction number R_eff<1 with closed borders) under various vaccine effectiveness (VE) scenarios and R₀ values. We ran two-year open-border simulations for two vaccine strategies: minimising R_eff and targeting high-risk groups. Findings: Targeting of high-risk groups will result in lower hospitalisations and deaths in most scenarios. Reaching the herd immunity threshold (HIT) with a vaccine of 90% VE against disease and 80% VE against infection requires at least 86•5% total population uptake for R₀=4•5 (with high vaccination coverage for 30-49-year-olds) and 98•1% uptake for R₀=6. In a two-year open-border scenario with 10 overseas cases daily and 90% total population vaccine uptake (including 0-15 year olds) with the same vaccine, the strategy of targeting high-risk groups is close to achieving HIT, with an estimated 11,400 total hospitalisations (peak 324 active and 36 new daily cases in hospitals), and 1,030 total deaths. Interpretation: Targeting high-risk groups for vaccination will result in fewer hospitalisations and deaths with open borders compared to targeting reduced transmission. With a highly effective vaccine and a high total uptake, opening borders will result in increasing cases, hospitalisations, and deaths. Other public health and social measures will still be required as part of an effective pandemic response. Funding: This project was funded by the Health Research Council [20/1018]. Research in context.

Postlicensure herpes zoster vaccine effectiveness: systematic review protocol

INTRODUCTION

Herpes zoster (HZ) and associated complications inflict substantial morbidity and associated healthcare and socioeconomic burdens. Current treatments are not fully effective, especially among the most vulnerable populations. Two HZ vaccines are available and are part of the national immunisation programmes in many countries. This review will evaluate the effectiveness of zoster vaccines against incident HZ and postherpetic neuralgia in adults 50 years and older.

METHODS AND ANALYSIS

The key information sources that will be searched include MEDLINE (Ovid), Embase (Ovid), Cochrane libraries and CINAHL. This search will consider postlicensure observational studies published in all languages between 2006 and 2020 that assessed the effectiveness of HZ/zoster vaccines in adults 50 years and older. The identification of studies will be complemented with the search of reference lists and citations, and contact with authors of papers to request missing or additional data, where required. Following the search, all identified citations will be collated, and duplicates will be removed. Titles and abstracts will then be screened by two independent reviewers for assessment against the inclusion criteria for the review. Selected studies will follow the process of critical appraisal, data extraction and data synthesis. Statistical analyses will be performed using a random-effect model.

ETHICS AND DISSEMINATION

Formal ethical approval is not required, as primary data will not be collected. The review will be disseminated in peer-reviewed publications and conference presentations.

Predicting Drug-Induced Liver Injury Using Convolutional Neural Network and Molecular Fingerprint-Embedded Features

As a critical issue in drug development and postmarketing safety surveillance, drug-induced liver injury (DILI) leads to failures in clinical trials as well as retractions of on-market approved drugs. Therefore, it is important to identify DILI compounds in the early-stages through in silico and in vivo studies. It is difficult using conventional safety testing methods, since the predictive power of most of the existing frameworks is insufficiently effective to address this pharmacological issue. In our study, we employ a natural language processing (NLP) inspired computational framework using convolutional neural networks and molecular fingerprint-embedded features. Our development set and independent test set have 1597 and 322 compounds, respectively. These samples were collected from previous studies and matched with established chemical databases for structural validity. Our study comes up with an average accuracy of 0.89, Matthews's correlation coefficient (MCC) of 0.80, and an AUC of 0.96. Our results show a significant improvement in the AUC values compared to the recent best model with a boost of 6.67%, from 0.90 to 0.96. Also, based on our findings, molecular fingerprint-embedded featurizer is an effective molecular representation for future biological and biochemical studies besides the application of classic molecular fingerprints.

High-content image generation for drug discovery using generative adversarial networks

Immense amount of high-content image data generated in drug discovery screening requires computationally driven automated analysis. Emergence of advanced machine learning algorithms, like deep learning models, has transformed the interpretation and analysis of imaging data. However, deep learning methods generally require large number of high-quality data samples, which could be limited during preclinical investigations. To address this issue, we propose a generative modeling based computational framework to synthesize images, which can be used for phenotypic profiling of perturbations induced by drug compounds. We investigated the use of three variants of Generative Adversarial Network (GAN) in our framework, viz., a basic Vanilla GAN, Deep Convolutional GAN (DCGAN) and Progressive GAN (ProGAN), and found DCGAN to be most efficient in generating realistic synthetic images. A pre-trained convolutional neural network (CNN) was used to extract features of both real and synthetic images, followed by a classification model trained on real and synthetic images. The quality of synthesized images was evaluated by comparing their feature distributions with that of real images. The DCGAN-based framework was applied to high-content image data from a drug screen to synthesize high-quality cellular images, which were used to augment the real image data. The augmented dataset was shown to yield better classification performance compared with that obtained using only real images. We also demonstrated the application of proposed method on the generation of bacterial images and computed feature distributions for bacterial images specific to different drug treatments. In summary, our results showed that the proposed DCGAN-based framework can be utilized to generate realistic synthetic high-content images, thus enabling the study of drug-induced effects on cells and bacteria.

iProDNA-CapsNet: identifying protein-DNA binding residues using capsule neural networks

BACKGROUND

Since protein-DNA interactions are highly essential to diverse biological events, accurately positioning the location of the DNA-binding residues is necessary. This biological issue, however, is currently a challenging task in the age of post-genomic where data on protein sequences have expanded very fast. In this study, we propose iProDNA-CapsNet - a new prediction model identifying protein-DNA binding residues using an ensemble of capsule neural networks (CapsNets) on position specific scoring matrix (PSMM) profiles. The use of CapsNets promises an innovative approach to determine the location of DNA-binding residues. In this study, the benchmark datasets introduced by Hu et al. (2017), i.e., PDNA-543 and PDNA-TEST, were used to train and evaluate the model, respectively. To fairly assess the model performance, comparative analysis between iProDNA-CapsNet and existing state-of-the-art methods was done.

RESULTS

Under the decision threshold corresponding to false positive rate (FPR) ≈ 5%, the accuracy, sensitivity, precision, and Matthews's correlation coefficient (MCC) of our model is increased by about 2.0%, 2.0%, 14.0%, and 5.0% with respect to TargetDNA (Hu et al., 2017) and 1.0%, 75.0%, 45.0%, and 77.0% with respect to BindN+ (Wang et al., 2010), respectively. With regards to other methods not reporting their threshold settings, iProDNA-CapsNet also shows a significant improvement in performance based on most of the evaluation metrics. Even with different patterns of change among the models, iProDNA-CapsNets remains to be the best model having top performance in most of the metrics, especially MCC which is boosted from about 8.0% to 220.0%.

CONCLUSIONS

According to all evaluation metrics under various decision thresholds, iProDNA-CapsNet shows better performance compared to the two current best models (BindN and TargetDNA). Our proposed approach also shows that CapsNet can potentially be used and adopted in other biological applications.

Classification of adaptor proteins using recurrent neural networks and PSSM profiles

BACKGROUND

Adaptor proteins are carrier proteins that play a crucial role in signal transduction. They commonly consist of several modular domains, each having its own binding activity and operating by forming complexes with other intracellular-signaling molecules. Many studies determined that the adaptor proteins had been implicated in a variety of human diseases. Therefore, creating a precise model to predict the function of adaptor proteins is one of the vital tasks in bioinformatics and computational biology. Few computational biology studies have been conducted to predict the protein functions, and in most of those studies, position specific scoring matrix (PSSM) profiles had been used as the features to be fed into the neural networks. However, the neural networks could not reach the optimal result because the sequential information in PSSMs has been lost. This study proposes an innovative approach by incorporating recurrent neural networks (RNNs) and PSSM profiles to resolve this problem.

RESULTS

Compared to other state-of-the-art methods which had been applied successfully in other problems, our method achieves enhancement in all of the common measurement metrics. The area under the receiver operating characteristic curve (AUC) metric in prediction of adaptor proteins in the cross-validation and independent datasets are 0.893 and 0.853, respectively.

CONCLUSIONS

This study opens a research path that can promote the use of RNNs and PSSM profiles in bioinformatics and computational biology. Our approach is reproducible by scientists that aim to improve the performance results of different protein function prediction problems. Our source code and datasets are available at https://github.com/ngphubinh/adaptors.

iEnhancer-ECNN: identifying enhancers and their strength using ensembles of convolutional neural networks

BACKGROUND

Enhancers are non-coding DNA fragments which are crucial in gene regulation (e.g. transcription and translation). Having high locational variation and free scattering in 98% of non-encoding genomes, enhancer identification is, therefore, more complicated than other genetic factors. To address this biological issue, several in silico studies have been done to identify and classify enhancer sequences among a myriad of DNA sequences using computational advances. Although recent studies have come up with improved performance, shortfalls in these learning models still remain. To overcome limitations of existing learning models, we introduce iEnhancer-ECNN, an efficient prediction framework using one-hot encoding and k-mers for data transformation and ensembles of convolutional neural networks for model construction, to identify enhancers and classify their strength. The benchmark dataset from Liu et al.'s study was used to develop and evaluate the ensemble models. A comparative analysis between iEnhancer-ECNN and existing state-of-the-art methods was done to fairly assess the model performance.

RESULTS

Our experimental results demonstrates that iEnhancer-ECNN has better performance compared to other state-of-the-art methods using the same dataset. The accuracy of the ensemble model for enhancer identification (layer 1) and enhancer classification (layer 2) are 0.769 and 0.678, respectively. Compared to other related studies, improvements in the Area Under the Receiver Operating Characteristic Curve (AUC), sensitivity, and Matthews's correlation coefficient (MCC) of our models are remarkable, especially for the model of layer 2 with about 11.0%, 46.5%, and 65.0%, respectively.

CONCLUSIONS

iEnhancer-ECNN outperforms other previously proposed methods with significant improvement in most of the evaluation metrics. Strong growths in the MCC of both layers are highly meaningful in assuring the stability of our models.

Predicting the onset of type 2 diabetes using wide and deep learning with electronic health records

OBJECTIVE

Diabetes is responsible for considerable morbidity, healthcare utilisation and mortality in both developed and developing countries. Currently, methods of treating diabetes are inadequate and costly so prevention becomes an important step in reducing the burden of diabetes and its complications. Electronic health records (EHRs) for each individual or a population have become important tools in understanding developing trends of diseases. Using EHRs to predict the onset of diabetes could improve the quality and efficiency of medical care. In this paper, we apply a wide and deep learning model that combines the strength of a generalised linear model with various features and a deep feed-forward neural network to improve the prediction of the onset of type 2 diabetes mellitus (T2DM).

MATERIALS AND METHODS

The proposed method was implemented by training various models into a logistic loss function using a stochastic gradient descent. We applied this model using public hospital record data provided by the Practice Fusion EHRs for the United States population. The dataset consists of de-identified electronic health records for 9948 patients, of which 1904 have been diagnosed with T2DM. Prediction of diabetes in 2012 was based on data obtained from previous years (2009-2011). The imbalance class of the model was handled by Synthetic Minority Oversampling Technique (SMOTE) for each cross-validation training fold to analyse the performance when synthetic examples for the minority class are created. We used SMOTE of 150 and 300 percent, in which 300 percent means that three new synthetic instances are created for each minority class instance. This results in the approximated diabetes:non-diabetes distributions in the training set of 1:2 and 1:1, respectively.

RESULTS

Our final ensemble model not using SMOTE obtained an accuracy of 84.28%, area under the receiver operating characteristic curve (AUC) of 84.13%, sensitivity of 31.17% and specificity of 96.85%. Using SMOTE of 150 and 300 percent did not improve AUC (83.33% and 82.12%, respectively) but increased sensitivity (49.40% and 71.57%, respectively) with a moderate decrease in specificity (90.16% and 76.59%, respectively).

DISCUSSION AND CONCLUSIONS

Our algorithm has further optimised the prediction of diabetes onset using a novel state-of-the-art machine learning algorithm: the wide and deep learning neural network architecture.

Superpixel-based segmentation of muscle fibers in multi-channel microscopy

BACKGROUND

Confetti fluorescence and other multi-color genetic labelling strategies are useful for observing stem cell regeneration and for other problems of cell lineage tracing. One difficulty of such strategies is segmenting the cell boundaries, which is a very different problem from segmenting color images from the real world. This paper addresses the difficulties and presents a superpixel-based framework for segmentation of regenerated muscle fibers in mice.

RESULTS

We propose to integrate an edge detector into a superpixel algorithm and customize the method for multi-channel images. The enhanced superpixel method outperforms the original and another advanced superpixel algorithm in terms of both boundary recall and under-segmentation error. Our framework was applied to cross-section and lateral section images of regenerated muscle fibers from confetti-fluorescent mice. Compared with "ground-truth" segmentations, our framework yielded median Dice similarity coefficients of 0.92 and higher.

CONCLUSION

Our segmentation framework is flexible and provides very good segmentations of multi-color muscle fibers. We anticipate our methods will be useful for segmenting a variety of tissues in confetti fluorecent mice and in mice with similar multi-color labels.

Hand gesture guided robot-assisted surgery based on a direct augmented reality interface

Radiofrequency (RF) ablation is a good alternative to hepatic resection for treatment of liver tumors. However, accurate needle insertion requires precise hand-eye coordination and is also affected by the difficulty of RF needle navigation. This paper proposes a cooperative surgical robot system, guided by hand gestures and supported by an augmented reality (AR)-based surgical field, for robot-assisted percutaneous treatment. It establishes a robot-assisted natural AR guidance mechanism that incorporates the advantages of the following three aspects: AR visual guidance information, surgeon's experiences and accuracy of robotic surgery. A projector-based AR environment is directly overlaid on a patient to display preoperative and intraoperative information, while a mobile surgical robot system implements specified RF needle insertion plans. Natural hand gestures are used as an intuitive and robust method to interact with both the AR system and surgical robot. The proposed system was evaluated on a mannequin model. Experimental results demonstrated that hand gesture guidance was able to effectively guide the surgical robot, and the robot-assisted implementation was found to improve the accuracy of needle insertion. This human-robot cooperative mechanism is a promising approach for precise transcutaneous ablation therapy.

An efficient compression scheme for 4-D medical images using hierarchical vector quantization and motion compensation

This paper proposes an efficient compression scheme for compressing time-varying medical volumetric data. The scheme uses 3-D motion estimation to create a homogenous preprocessed data to be compressed by a 3-D image compression algorithm using hierarchical vector quantization. A new block distortion measure, called variance of residual (VOR), and three 3-D fast block matching algorithms are used to improve the motion estimation process in term of speed and data fidelity. The 3-D image compression process involves the application of two different encoding techniques based on the homogeneity of input data. Our method can achieve a higher fidelity and faster decompression time compared to other lossy compression methods producing similar compression ratios. The combination of 3-D motion estimation using VOR and hierarchical vector quantization contributes to the good performance.

Ligation of integrin alpha 3beta 1 by laminin 5 at the wound edge activates Rho-dependent adhesion of leading keratinocytes on collagen

Wounding of the epidermis signals the transition of keratinocytes from quiescent anchorage on endogenous basement membrane laminin 5 to migration on exposed dermal collagen. In this study, we attempt to characterize activation signals that transform quiescent keratinocytes into migratory leading cells at the wound edge. Previously, we reported that adhesion and spreading on collagen via integrin alpha(2)beta(1) by cultured human foreskin keratinocytes (HFKs) requires RhoGTP, a regulator of actin stress fibers. In contrast, adhesion and spreading on laminin 5 requires integrins alpha(3)beta(1) and alpha(6)beta(4) and is dependent on phosphoinositide 3-hydroxykinase (Nguyen, B. P., Gil, S. G., and Carter, W. G. (2000) J. Biol. Chem. 275, 31896-31907). Here, we report that quiescent HFKs do not adhere to collagen but adhere and spread on laminin 5. By using collagen adhesion as one criterion for conversion to a "leading wound cell," we found that activation of collagen adhesion requires elevation of RhoGTP. Adhesion of quiescent HFKs to laminin 5 via integrin alpha(3)beta(1) and alpha(6)beta(4) is sufficient to increase levels of RhoGTP required for adhesion and spreading on collagen. Consistently, adhesion of quiescent HFKs to laminin 5, but not collagen, also promotes expression of the precursor form of laminin 5, a characteristic of leading keratinocytes in the epidermal outgrowth. We suggest that wounding of quiescent epidermis initiates adhesion and spreading of keratinocytes at the wound edge on endogenous basement membrane laminin 5 via alpha(3)beta(1) and alpha(6)beta(4) in a Rho-independent mechanism. Spreading on endogenous laminin 5 via alpha(3)beta(1) is necessary but not sufficient to elevate expression of precursor laminin 5 and RhoGTP, allowing for subsequent collagen adhesion via alpha(2)beta(1), all characteristics of leading keratinocytes in the epidermal outgrowth.

Deposition of laminin 5 by keratinocytes regulates integrin adhesion and signaling

Deposition of laminin 5 over exposed dermal collagen in epidermal wounds is an early event in repair of the basement membrane. We report that deposition of laminin 5 onto collagen switches adhesion and signaling from collagen-dependent to laminin 5-dependent. Ligation of laminin 5 by integrin alpha(6)beta(4) activates phosphoinositide 3-OH-kinase (PI3K) signaling. This activation allows for adhesion and spreading via integrin alpha(3)beta(1) on laminin 5 independent of RhoGTPase, a regulator of actin stress fibers. In contrast, adhesion and spreading on collagen via alpha(2)beta(1) is Rho-dependent and is inhibited by toxin B, a Rho inhibitor. Deposition of laminin 5 and ligation of alpha(6)beta(4) increases PI3K-dependent production of phosphoinositide di- and triphosphates, PI3K activity, and phosphorylation of downstream target protein c-Jun NH(2)-terminal kinase. Conversely, blocking laminin 5-deposition with brefeldin A, an inhibitor of vesicle transport, or with anti-laminin 5 monoclonal antibodies abolishes the PI3K-dependent spreading mediated by alpha(3)beta(1) and phosphorylation of c-Jun NH(2)-terminal kinase. Studies with keratinocytes lacking alpha(6)beta(4) or laminin 5 confirm that deposition of laminin 5 and ligation by alpha(6)beta(4) are required for PI3K-dependent spreading via alpha(3)beta(1). We suggest that deposition of laminin 5 onto the collagen substratum, as in wound repair, enables human foreskin keratinocytes to interact via alpha(6)beta(4) and to switch from a RhoGTPase-dependent adhesion on collagen to a PI3K-dependent adhesion and spreading mediated by integrin alpha(3)beta(1) on laminin 5.

Deposition of laminin 5 in epidermal wounds regulates integrin signaling and adhesion

Adhesion of keratinocytes in a wound outgrowth to laminin 5 in the basement membrane via integrins alpha6beta4 and alpha3beta1 is distinct from adhesion to dermal collagen via alpha2beta1 or to fibronectin via alpha5beta1. Leading cells in the outgrowth are distinguished from following keratinocytes by deposition of laminin 5, failure to communicate via gap junctions and sensitivity to toxin B, an inhibitor of RhoGTPase. Laminin 5 deposited by leading keratinocytes onto dermal collagen dominates over dermal ligands and changes the cell signals required for adhesion from collagen-dependent to laminin-5-dependent. Thus, deposition of laminin 5 can instruct keratinocytes to switch from an activated phenotype to a quiescent and integrated epithelial phenotype.

Cellular interaction of integrin alpha3beta1 with laminin 5 promotes gap junctional communication

Wounding of skin activates epidermal cell migration over exposed dermal collagen and fibronectin and over laminin 5 secreted into the provisional basement membrane. Gap junctional intercellular communication (GJIC) has been proposed to integrate the individual motile cells into a synchronized colony. We found that outgrowths of human keratinocytes in wounds or epibole cultures display parallel changes in the expression of laminin 5, integrin alpha3beta1, E-cadherin, and the gap junctional protein connexin 43. Adhesion of keratinocytes on laminin 5, collagen, and fibronectin was found to differentially regulate GJIC. When keratinocytes were adhered on laminin 5, both structural (assembly of connexin 43 in gap junctions) and functional (dye transfer) assays showed a two- to threefold increase compared with collagen and five- to eightfold over fibronectin. Based on studies with immobilized integrin antibody and integrin-transfected Chinese hamster ovary cells, the interaction of integrin alpha3beta1 with laminin 5 was sufficient to promote GJIC. Mapping of intermediate steps in the pathway linking alpha3beta1-laminin 5 interactions to GJIC indicated that protein trafficking and Rho signaling were both required. We suggest that adhesion of epithelial cells to laminin 5 in the basement membrane via alpha3beta1 promotes GJIC that integrates individual cells into synchronized epiboles.

Patients' perceptions compared with objective ratings of asthma severity

BACKGROUND

Optimal asthma management requires accurate assessment of asthma severity.

OBJECTIVE

To compare patients' perceptions of their asthma severity with that obtained by using the guidelines published by the National Asthma Education and Prevention Program's Expert Panel and with functional impairment measured by spirometry and numeric criteria of the American Thoracic Society.

METHODS

We enrolled 323 patients age 18 to 50 years who were members of the Kaiser Foundation Health Plan for > or = 1 year in a randomized control trial of an asthma education program. Each had a confirmed diagnosis of bronchial asthma and had been receiving antiasthma medication for > or = 1 year. Patients rated the severity of their asthma. Office spirometry was performed, and, using the Mini-Wright peak flow meter, patients kept 2-week diaries of at-home recordings of morning and evening peak expiratory flow rates.

RESULTS

A statistically significant association was noted between patients' perceptions of asthma severity and both medication severity rating (P < .001) and diurnal variation rating (P = .003) and evening peak expiratory flow rate percentage (P = .019). In comparison with a severity composite based on criteria of the National Asthma Education Program, 54% of patients accurately estimated asthma severity, 27% overestimated, and 20% underestimated severity.

CONCLUSION

A clinically significant proportion of asthmatic patients substantially underestimate disease severity and thereby may be at risk of increased mortality or morbidity.

Intravenous fenoldopam versus sodium nitroprusside in patients with severe hypertension

In an open-label study, we compared the efficacy and safety of intravenous infusion of fenoldopam mesylate with that of sodium nitroprusside in patients with severe hypertension or in hypertensive crisis. Both antihypertensive medications were infused at a maximal dose increment of 0.2 microgram/kg/min (fenoldopam) and 1 microgram/kg/min (nitroprusside), with a maximal infusion rate of 1.5 micrograms/kg/min fenoldopam mesylate or 8 micrograms/kg/min sodium nitroprusside. Once the desired reduction in diastolic blood pressure was achieved (less than 110 mm Hg if initial diastolic blood pressure was 120-149 mm Hg, or by at least 40 mm Hg if initial diastolic blood pressure was 150-190 mm Hg), the maximal infusion rate used was maintained for at least 1 hour, and then, the infusion was slowed gradually over 2 hours. After the infusion treatment, patients remained in the hospital for 2 days of follow-up. Both antihypertensive agents successfully controlled the blood pressure in all the patients by the end of the maintenance periods. Between the baseline and the end of the maintenance period, analysis of variance showed that the changes in the variables induced by fenoldopam mesylate did not differ significantly from those induced by sodium nitroprusside. The incidence of side effects listed were similar in both groups of patients. The detection of toxic levels of thiocyanate in two patients treated with nitroprusside, however, shows that fenoldopam might be preferable for the control of a hypertensive crisis or severe hypertension in patients with decreased renal function.

Macronodular lesions associated with Staphylococcus aureus bacteremia. A condition that resembles a 'lymphocutaneous' syndrome

A 54-year-old man, diagnosed as having Wegener's granulomatosis and treated with a regimen of cyclophosphamide and prednisone and hemodialysis, was found to have Staphylococcus aureus in blood, urine, and pus that were removed from the infected area. He had unusual macronodular lesions of the skin that resembled the lymphocutaneous syndrome. These lesions resolved with antibiotic therapy. To our knowledge, this condition has not been described previously.