CFNCM: Collaborative filtering neighborhood-based model for predicting miRNA-disease associations

MicroRNAs have a significant role in the emergence of various human disorders. Consequently, it is essential to understand the existing interactions between miRNAs and diseases, as this will help scientists better study and comprehend the diseases' biological mechanisms. Findings can be employed as biomarkers or drug targets to advance the detection, diagnosis, and treatment of complex human disorders by foretelling possible disease-related miRNAs. This study proposed a computational model for predicting potential miRNA-disease associations called the Collaborative Filtering Neighborhood-based Classification Model (CFNCM), in light of the shortcomings of conventional and biological experiments, which are expensive and time-consuming. The model generated integrated miRNA and disease similarity matrices using the validated associations and miRNA and disease similarity information and used them as the input features for CFNCM. To produce class labels, we first determined the association scores for brand-new pairs using user-based collaborative filtering. With zero as the threshold, the associations with scores >0 were labelled 1, indicating a potential positive association, otherwise, it is marked as 0. Then, we developed classification models using various machine-learning algorithms. By comparison, we discovered that the support vector machine (SVM) produced the best AUC of 0.96 with 10-fold cross-validation through the GridSearchCV technique for identifying optimal parameter values. In addition, the models were evaluated and verified by analyzing the top 50 breast and lung neoplasms-related miRNAs, of which 46 and 47 associations were verified in two authoritative databases, dbDEMC and miR2Disease.

A genome-wide association study of serum proteins reveals shared loci with common diseases

With the growing number of genetic association studies, the genotype-phenotype atlas has become increasingly more complex, yet the functional consequences of most disease associated alleles is not understood. The measurement of protein level variation in solid tissues and biofluids integrated with genetic variants offers a path to deeper functional insights. Here we present a large-scale proteogenomic study in 5,368 individuals, revealing 4,035 independent associations between genetic variants and 2,091 serum proteins, of which 36% are previously unreported. The majority of both cis- and trans-acting genetic signals are unique for a single protein, although our results also highlight numerous highly pleiotropic genetic effects on protein levels and demonstrate that a protein's genetic association profile reflects certain characteristics of the protein, including its location in protein networks, tissue specificity and intolerance to loss of function mutations. Integrating protein measurements with deep phenotyping of the cohort, we observe substantial enrichment of phenotype associations for serum proteins regulated by established GWAS loci, and offer new insights into the interplay between genetics, serum protein levels and complex disease.

Coding and regulatory variants are associated with serum protein levels and disease

Circulating proteins can be used to diagnose and predict disease-related outcomes. A deep serum proteome survey recently revealed close associations between serum protein networks and common disease. In the current study, 54,469 low-frequency and common exome-array variants were compared to 4782 protein measurements in the serum of 5343 individuals from the AGES Reykjavik cohort. This analysis identifies a large number of serum proteins with genetic signatures overlapping those of many diseases. More specifically, using a study-wide significance threshold, we find that 2021 independent exome array variants are associated with serum levels of 1942 proteins. These variants reside in genetic loci shared by hundreds of complex disease traits, highlighting serum proteins' emerging role as biomarkers and potential causative agents of a wide range of diseases.

Questioning Established Theories and Treatment Methods Related to Iron and Other Metal Metabolic Changes, Affecting All Major Diseases and Billions of Patients

The medical and scientific literature is dominated by highly cited historical theories and findings [...].

Characteristics of Disease-Specific and Generic Diagnostic Pitfalls: A Qualitative Study

IMPORTANCE

Progress in understanding and preventing diagnostic errors has been modest. New approaches are needed to help clinicians anticipate and prevent such errors. Delineating recurring diagnostic pitfalls holds potential for conceptual and practical ways for improvement.

OBJECTIVES

To develop the construct and collect examples of "diagnostic pitfalls," defined as clinical situations and scenarios vulnerable to errors that may lead to missed, delayed, or wrong diagnoses.

DESIGN, SETTING, AND PARTICIPANTS

This qualitative study used data from January 1, 2004, to December 31, 2016, from retrospective analysis of diagnosis-related patient safety incident reports, closed malpractice claims, and ambulatory morbidity and mortality conferences, as well as specialty focus groups. Data analyses were conducted between January 1, 2017, and December 31, 2019.

MAIN OUTCOMES AND MEASURES

From each data source, potential diagnostic error cases were identified, and the following information was extracted: erroneous and correct diagnoses, presenting signs and symptoms, and areas of breakdowns in the diagnostic process (using Diagnosis Error Evaluation and Research and Reliable Diagnosis Challenges taxonomies). From this compilation, examples were collected of disease-specific pitfalls; this list was used to conduct a qualitative analysis of emerging themes to derive a generic taxonomy of diagnostic pitfalls.

RESULTS

A total of 836 relevant cases were identified among 4325 patient safety incident reports, 403 closed malpractice claims, 24 ambulatory morbidity and mortality conferences, and 355 focus groups responses. From these, 661 disease-specific diagnostic pitfalls were identified. A qualitative review of these disease-specific pitfalls identified 21 generic diagnostic pitfalls categories, which included mistaking one disease for another disease (eg, aortic dissection is misdiagnosed as acute myocardial infarction), failure to appreciate test result limitations, and atypical disease presentations.

CONCLUSIONS AND RELEVANCE

Recurring types of pitfalls were identified and collected from diagnostic error cases. Clinicians could benefit from knowledge of both disease-specific and generic cross-cutting pitfalls. Study findings can potentially inform educational and quality improvement efforts to anticipate and prevent future errors.

Determine independent gut microbiota-diseases association by eliminating the effects of human lifestyle factors

Lifestyle and physiological variables on human disease risk have been revealed to be mediated by gut microbiota. Low concordance between case-control studies for detecting disease-associated microbe existed due to limited sample size and population-wide bias in lifestyle and physiological variables. To infer gut microbiota-disease associations accurately, we propose to build machine learning models by including both human variables and gut microbiota. When the model's performance with both gut microbiota and human variables is better than the model with just human variables, the independent gut microbiota -disease associations will be confirmed. By building models on the American Gut Project dataset, we found that gut microbiota showed distinct association strengths with different diseases. Adding gut microbiota into human variables enhanced the classification performance of IBD significantly; independent associations between occurrence information of gut microbiota and irritable bowel syndrome, C. difficile infection, and unhealthy status were found; adding gut microbiota showed no improvement on models' performance for diabetes, small intestinal bacterial overgrowth, lactose intolerance, cardiovascular disease. Our results suggested that although gut microbiota was reported to be associated with many diseases, a considerable proportion of these associations may be very weak. We proposed a list of microbes as biomarkers to classify IBD and unhealthy status. Further functional investigations of these microbes will improve understanding of the molecular mechanism of human diseases.

Cross-sectional study of the ambulance transport between healthcare facilities with medical support via telemedicine: Easy, effective, and safe tool

Background

Feasibility and safety of ambulance transport between healthcare facilities with medical support exclusively via telemedicine are unknown.

Methods

This was a retrospective study with a single telemedicine center reference for satellite emergency departments of the same hospital. The study population was all critically ill patients admitted to one of the peripheral units from November 2016 to May 2020 and who needed to be transferred to the main building. Telemedicine-assisted transportation was performed by an emergency specialist. The inclusion criteria included patients above the age of 15 and initial stabilization performed at the emergency department. Unstable, intubated, ST-elevation myocardial infarction and acute stroke patients were excluded. There was a double-check of safety conditions by the nurse and the remote doctor before the ambulance departure. The primary endpoint was the number of telemedicine-guided interventions during transport.

Results

2840 patients were enrolled. The population was predominantly male (53.2%) with a median age of 60 years. Sepsis was the most prevalent diagnosis in 28% of patients, followed by acute coronary syndromes (8.5%), arrhythmia (6.7%), venous thromboembolism (6.1%), stroke (6.1%), acute abdomen (3.6%), respiratory distress (3.3%), and heart failure (2.5%). Only 22 (0.8%) patients required telemedicine-assisted support during transport. Administration of oxygen therapy and analgesics were the most common recommendations made by telemedicine emergency physicians. There were no communication problems in the telemedicine-assisted group.

Conclusions

Telemedicine-assisted ambulance transportation between healthcare facilities of stabilized critically ill patients may be an option instead of an onboard physician. The frequency of clinical support requests by telemedicine is minimal, and most evaluations are of low complexity and easily and safely performed by trained nurses.

DeepCNV: a deep learning approach for authenticating copy number variations

Copy number variations (CNVs) are an important class of variations contributing to the pathogenesis of many disease phenotypes. Detecting CNVs from genomic data remains difficult, and the most currently applied methods suffer from an unacceptably high false positive rate. A common practice is to have human experts manually review original CNV calls for filtering false positives before further downstream analysis or experimental validation. Here, we propose DeepCNV, a deep learning-based tool, intended to replace human experts when validating CNV calls, focusing on the calls made by one of the most accurate CNV callers, PennCNV. The sophistication of the deep neural network algorithm is enriched with over 10 000 expert-scored samples that are split into training and testing sets. Variant confidence, especially for CNVs, is a main roadblock impeding the progress of linking CNVs with the disease. We show that DeepCNV adds to the confidence of the CNV calls with an optimal area under the receiver operating characteristic curve of 0.909, exceeding other machine learning methods. The superiority of DeepCNV was also benchmarked and confirmed using an experimental wet-lab validation dataset. We conclude that the improvement obtained by DeepCNV results in significantly fewer false positive results and failures to replicate the CNV association results.

Deep learning in systems medicine

Systems medicine (SM) has emerged as a powerful tool for studying the human body at the systems level with the aim of improving our understanding, prevention and treatment of complex diseases. Being able to automatically extract relevant features needed for a given task from high-dimensional, heterogeneous data, deep learning (DL) holds great promise in this endeavour. This review paper addresses the main developments of DL algorithms and a set of general topics where DL is decisive, namely, within the SM landscape. It discusses how DL can be applied to SM with an emphasis on the applications to predictive, preventive and precision medicine. Several key challenges have been highlighted including delivering clinical impact and improving interpretability. We used some prototypical examples to highlight the relevance and significance of the adoption of DL in SM, one of them is involving the creation of a model for personalized Parkinson's disease. The review offers valuable insights and informs the research in DL and SM.

Association between a complex community intervention and quality of health extension workers' performance to correctly classify common childhood illnesses in four regions of Ethiopia

Background

Due to low care utilization, a complex intervention was done for two years to optimize the Ethiopian Health Extension Program. Improved quality of the integrated community case management services was an intermediate outcome of this intervention through community education and mobilization, capacity building of health workers, and strengthening of district ownership and accountability of sick child services. We evaluated the association between the intervention and the health extension workers’ ability to correctly classify common childhood illnesses in four regions of Ethiopia.

Methods

Baseline and endline assessments were done in 2016 and 2018 in intervention and comparison areas in four regions of Ethiopia. Ill children aged 2 to 59 months were mobilized to visit health posts for an assessment that was followed by re-examination. We analyzed sensitivity, specificity, and difference-in-difference of correct classification with multilevel mixed logistic regression in intervention and comparison areas at baseline and endline.

Results

Health extensions workers’ consultations with ill children were observed in intervention (n = 710) and comparison areas (n = 615). At baseline, re-examination of the children showed that in intervention areas, health extension workers’ sensitivity for fever or malaria was 54%, 68% for respiratory infections, 90% for diarrheal diseases, and 34% for malnutrition. At endline, it was 40% for fever or malaria, 49% for respiratory infections, 85% for diarrheal diseases, and 48% for malnutrition. Specificity was higher (89–100%) for all childhood illnesses. Difference-in-differences was 6% for correct classification of fever or malaria [aOR = 1.45 95% CI: 0.81–2.60], 4% for respiratory tract infection [aOR = 1.49 95% CI: 0.81–2.74], and 5% for diarrheal diseases [aOR = 1.74 95% CI: 0.77–3.92].

Conclusion

This study revealed that the Optimization of Health Extension Program intervention, which included training, supportive supervision, and performance reviews of health extension workers, was not associated with an improved classification of childhood illnesses by these Ethiopian primary health care workers.

Trial registration

ISRCTN12040912, http://www.isrctn.com/ISRCTN12040912.

SynTEG: a framework for temporal structured electronic health data simulation

OBJECTIVE

Simulating electronic health record data offers an opportunity to resolve the tension between data sharing and patient privacy. Recent techniques based on generative adversarial networks have shown promise but neglect the temporal aspect of healthcare. We introduce a generative framework for simulating the trajectory of patients' diagnoses and measures to evaluate utility and privacy.

MATERIALS AND METHODS

The framework simulates date-stamped diagnosis sequences based on a 2-stage process that 1) sequentially extracts temporal patterns from clinical visits and 2) generates synthetic data conditioned on the learned patterns. We designed 3 utility measures to characterize the extent to which the framework maintains feature correlations and temporal patterns in clinical events. We evaluated the framework with billing codes, represented as phenome-wide association study codes (phecodes), from over 500 000 Vanderbilt University Medical Center electronic health records. We further assessed the privacy risks based on membership inference and attribute disclosure attacks.

RESULTS

The simulated temporal sequences exhibited similar characteristics to real sequences on the utility measures. Notably, diagnosis prediction models based on real versus synthetic temporal data exhibited an average relative difference in area under the ROC curve of 1.6% with standard deviation of 3.8% for 1276 phecodes. Additionally, the relative difference in the mean occurrence age and time between visits were 4.9% and 4.2%, respectively. The privacy risks in synthetic data, with respect to the membership and attribute inference were negligible.

CONCLUSION

This investigation indicates that temporal diagnosis code sequences can be simulated in a manner that provides utility and respects privacy.

Generative transfer learning for measuring plausibility of EHR diagnosis records

OBJECTIVE

Due to a complex set of processes involved with the recording of health information in the Electronic Health Records (EHRs), the truthfulness of EHR diagnosis records is questionable. We present a computational approach to estimate the probability that a single diagnosis record in the EHR reflects the true disease.

MATERIALS AND METHODS

Using EHR data on 18 diseases from the Mass General Brigham (MGB) Biobank, we develop generative classifiers on a small set of disease-agnostic features from EHRs that aim to represent Patients, pRoviders, and their Interactions within the healthcare SysteM (PRISM features).

RESULTS

We demonstrate that PRISM features and the generative PRISM classifiers are potent for estimating disease probabilities and exhibit generalizable and transferable distributional characteristics across diseases and patient populations. The joint probabilities we learn about diseases through the PRISM features via PRISM generative models are transferable and generalizable to multiple diseases.

DISCUSSION

The Generative Transfer Learning (GTL) approach with PRISM classifiers enables the scalable validation of computable phenotypes in EHRs without the need for domain-specific knowledge about specific disease processes.

CONCLUSION

Probabilities computed from the generative PRISM classifier can enhance and accelerate applied Machine Learning research and discoveries with EHR data.

Climate and categories: Two key elements for understanding the Mesoamerican hot-cold classification of illnesses and medicinal plants

ETHNOPHARMACOLOGICAL RELEVANCE

The concepts of health and illness, and their causes, are fundamental for understanding medicinal plant choice and use by traditional people. The hot-cold system is widespread in Mesoamerican traditional medicine and guides many therapeutic decisions.

AIM OF THE STUDY

This study explores a hypothesis that climate influences the hot-cold classification of illnesses and medicinal plants, and the perception of hazard of illnesses. In addition, we examine the classification categories within the system used in different regions of Mexico.

MATERIALS AND METHODS

Studies from Mexico with quantitative and qualitative data on the hot-cold properties of medicinal plants and ailments were reviewed. The information was organized and then related to the climate type of the study areas.

RESULTS

In temperate climates, most diseases were considered cold, and hot medicinal plants were dominant. Conversely, in warm-tropical climates, hot diseases dominated, and the majority of medicinal plants were cold; however, this evidence was weaker. The perception of hazard was congruent with the number of illnesses for temperate climates. There were additional classification categories within the hot-cold system for diseases and medicinal plants, and they were expressed in different terms in Spanish, English, and indigenous languages. Although similar terms and categories were used in the classification of diseases and medicinal plants, they can differ conceptually and vary between places and cultures. Publications are sometimes unclear if the terms used are emic or etic. The basic principle of using plants with the opposite property of the disease does not always apply strictly.

CONCLUSIONS

Climate appears to influence the hot-cold classification of diseases and medicinal plants in Mexico, and the system is not strictly dual. Improved knowledge of the hot-cold system is necessary to understand Mesoamerican medicinal plant use and culture.

A typology of clinical conditions

In the philosophy of medicine, great attention has been paid to defining disease, yet less attention has been paid to the classification of clinical conditions. These include conditions that look like diseases but are not; conditions that are diseases but that (currently) have no diagnostic criteria; and other types, including those relating to risk for disease. I present a typology of clinical conditions by examining factors important for characterizing clinical conditions. By attending to the types of clinical conditions possible on the basis of these key factors (symptomaticity, dysfunction, and the meeting of diagnostic criteria), I draw attention to how diseases and other clinical conditions as currently classified can be better categorized, highlighting the issues pertaining to certain typology categories. Through detailed analysis of a wide variety of clinical examples, including Alzheimer disease as a test case, I show how nosology, research, and decisions about diagnostic criteria should include normative as well as naturalistically describable factors.

Similar Disease Prediction With Heterogeneous Disease Information Networks

Studying the similarity of diseases can help us to explore the pathological characteristics of complex diseases, and help provide reliable reference information for inferring the relationship between new diseases and known diseases, so as to develop effective treatment plans. To obtain the similarity of the disease, most previous methods either use a single similarity metric such as semantic score, functional score from single data source, or utilize weighting coefficients to simply combine multiple metrics with different dimensions. In this paper, we proposes a method to predict the similarity of diseases by node representation learning. We first integrate the semantic score and topological score between diseases by combining multiple data sources. Then for each disease, its integrated scores with all other diseases are utilized to map it into a vector of the same spatial dimension, and the vectors are used to measure and comprehensively analyze the similarity between diseases. Lastly, we conduct comparative experiment based on benchmark set and other disease nodes outside the benchmark set. Using the statistics such as average, variance, and coefficient of variation in the benchmark set to evaluate multiple methods demonstrates the effectiveness of our approach in the prediction of similar diseases.

It is time to apply biclustering: a comprehensive review of biclustering applications in biological and biomedical data

Biclustering is a powerful data mining technique that allows clustering of rows and columns, simultaneously, in a matrix-format data set. It was first applied to gene expression data in 2000, aiming to identify co-expressed genes under a subset of all the conditions/samples. During the past 17 years, tens of biclustering algorithms and tools have been developed to enhance the ability to make sense out of large data sets generated in the wake of high-throughput omics technologies. These algorithms and tools have been applied to a wide variety of data types, including but not limited to, genomes, transcriptomes, exomes, epigenomes, phenomes and pharmacogenomes. However, there is still a considerable gap between biclustering methodology development and comprehensive data interpretation, mainly because of the lack of knowledge for the selection of appropriate biclustering tools and further supporting computational techniques in specific studies. Here, we first deliver a brief introduction to the existing biclustering algorithms and tools in public domain, and then systematically summarize the basic applications of biclustering for biological data and more advanced applications of biclustering for biomedical data. This review will assist researchers to effectively analyze their big data and generate valuable biological knowledge and novel insights with higher efficiency.

Machine learning and clinical epigenetics: a review of challenges for diagnosis and classification

BACKGROUND

Machine learning is a sub-field of artificial intelligence, which utilises large data sets to make predictions for future events. Although most algorithms used in machine learning were developed as far back as the 1950s, the advent of big data in combination with dramatically increased computing power has spurred renewed interest in this technology over the last two decades.

MAIN BODY

Within the medical field, machine learning is promising in the development of assistive clinical tools for detection of e.g. cancers and prediction of disease. Recent advances in deep learning technologies, a sub-discipline of machine learning that requires less user input but more data and processing power, has provided even greater promise in assisting physicians to achieve accurate diagnoses. Within the fields of genetics and its sub-field epigenetics, both prime examples of complex data, machine learning methods are on the rise, as the field of personalised medicine is aiming for treatment of the individual based on their genetic and epigenetic profiles.

CONCLUSION

We now have an ever-growing number of reported epigenetic alterations in disease, and this offers a chance to increase sensitivity and specificity of future diagnostics and therapies. Currently, there are limited studies using machine learning applied to epigenetics. They pertain to a wide variety of disease states and have used mostly supervised machine learning methods.

International Olympic Committee consensus statement: methods for recording and reporting of epidemiological data on injury and illness in sport 2020 (including STROBE Extension for Sport Injury and Illness Surveillance (STROBE-SIIS))

Injury and illness surveillance, and epidemiological studies, are fundamental elements of concerted efforts to protect the health of the athlete. To encourage consistency in the definitions and methodology used, and to enable data across studies to be compared, research groups have published 11 sport-specific or setting-specific consensus statements on sports injury (and, eventually, illness) epidemiology to date. Our objective was to further strengthen consistency in data collection, injury definitions and research reporting through an updated set of recommendations for sports injury and illness studies, including a new Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist extension. The IOC invited a working group of international experts to review relevant literature and provide recommendations. The procedure included an open online survey, several stages of text drafting and consultation by working groups and a 3-day consensus meeting in October 2019. This statement includes recommendations for data collection and research reporting covering key components: defining and classifying health problems; severity of health problems; capturing and reporting athlete exposure; expressing risk; burden of health problems; study population characteristics and data collection methods. Based on these, we also developed a new reporting guideline as a STROBE Extension-the STROBE Sports Injury and Illness Surveillance (STROBE-SIIS). The IOC encourages ongoing in- and out-of-competition surveillance programmes and studies to describe injury and illness trends and patterns, understand their causes and develop measures to protect the health of the athlete. Implementation of the methods outlined in this statement will advance consistency in data collection and research reporting.

Limited alignment of publicly competitive disease funding with disease burden in Japan

OBJECTIVE

The need to align investments in health research and development (R&D) with public health needs is one of the most important public health challenges in Japan. We examined the alignment of disease-specific publicly competitive R&D funding to the disease burden in the country.

METHODS

We analyzed publicly available data on competitive public funding for health in 2015 and 2016 and compared it to disability-adjusted life year (DALYs) in 2016, which were obtained from the Global Burden of Disease (GBD) 2017 study. Their alignment was assessed as a percentage distribution among 22 GBD disease groups. Funding was allocated to the 22 disease groups based on natural language processing, using textual information such as project title and abstract for each research project, while considering for the frequency of information.

RESULTS

Total publicly competitive funding in health R&D in 2015 and 2016 reached 344.1 billion JPY (about 3.0 billion USD) for 32,204 awarded projects. About 49.5% of the funding was classifiable for disease-specific projects. Five GDB disease groups were significantly and relatively well-funded compared to their contributions to Japan's DALY, including neglected tropical diseases and malaria (funding vs DALY = 1.7% vs 0.0%, p<0.01) and neoplasms (28.5% vs 19.2%, p<0.001). In contrast, four GDB disease groups were significantly under-funded, including cardiovascular diseases (8.0% vs 14.8%, p<0.001) and musculoskeletal disorders (1.0% vs 11.9%, p<0.001). These percentages do not include unclassifiable funding.

CONCLUSIONS

While caution is necessary as this study was not able to consider public in-house funding and the methodological uncertainties could not be ruled out, the analysis may provide a snapshot of the limited alignment between publicly competitive disease-specific funding and the disease burden in the country. The results call for greater management over the allocation of scarce resources on health R&D. DALYs will serve as a crucial, but not the only, consideration in aligning Japan's research priorities with the public health needs. In addition, the algorithms for natural language processing used in this study require continued efforts to improve accuracy.

Medical disorder, harm, and damage

Jerome Wakefield's harmful dysfunction analysis (HDA) of medical disorder is an influential hybrid of naturalist and normative theories. In order to conclude that a condition is a disorder, according to the HDA, one must determine both that it results from a failure of a physical or psychological mechanism to perform its natural function and that it is harmful. In a recent issue of this journal, I argued that the HDA entails implausible judgments about which disorders there are and how they are individuated. The same arguments apply to other views that incorporate a harm criterion. More recently, David G. Limbaugh has modified the HDA by providing a novel account of the way in which a disorder must be harmful. Here, I briefly review the relevant issues and then critically assess Limbaugh's account. I argue in the end that Limbaugh's revisions do not succeed in making accounts like the HDA more attractive.

R-Ensembler: A greedy rough set based ensemble attribute selection algorithm with kNN imputation for classification of medical data

BACKGROUND AND OBJECTIVE

Retrieving meaningful information from high dimensional dataset is an important and challenging task. Normally, medical dataset suffers from several issues such as curse of dimensionality problem, uncertainty, presence of missing values, non-relevant and redundant attributes, etc. Any machine learning technique applied on such data (without any preprocessing) by and large takes a considerable amount of computational time and may degrade the performance of the model.

METHODS

In this article, R-Ensembler, a parameter free greedy ensemble attribute selection method is proposed adopting the concept of rough set theory by using the attribute-class, attribute-significance and attribute-attribute relevance measures to select a subset of attributes which are most relevant, significant and non-redundant from a pool of different attribute subsets in order to predict the presence or absence of different diseases in medical dataset. The main role of the proposed ensembler is to combine multiple subsets of attributes produced by different rough set filters and to produce an optimal subset of attributes for subsequent classification task. A novel n number of set intersection method is also proposed to reduce the biasness during the time of attribute selection process. Before selecting the minimal attribute set from a given data by the proposed R-Ensembler method, the dataset is preprocessed by the k nearest neighbour (kNN) imputation method for missing value treatment.

RESULTS

Experiments are carried out on seven benchmark medical datasets collected from University of California at Irvine (UCI) repository. The performance of the proposed ensemble method is compared with five state-of-the-art attribute selection algorithms, results of which are measured using three benchmark classifiers viz., Naïve Bayes, decision trees and random forest. Experimental results clearly justify the superiority of the proposed R-Ensembler method over other attribute selection algorithms. Results of paired t-test performed on average accuracies produced by different classifiers simulated on the reduced data sets achieved by the proposed and counter part attribute selection methods confirm the statistical significance of the better reduced attribute subsets achieved by the proposed R-Ensembler method compared to others.

CONCLUSION

The proposed ensemble method turned out to be very effective for selecting high relevant, high significant and less redundant attributes from a pool of different subsets of attributes.

MultiSourcDSim: an integrated approach for exploring disease similarity

BACKGROUND

A collection of disease-associated data contributes to study the association between diseases. Discovering closely related diseases plays a crucial role in revealing their common pathogenic mechanisms. This might further imply treatment that can be appropriated from one disease to another. During the past decades, a number of approaches for calculating disease similarity have been developed. However, most of them are designed to take advantage of single or few data sources, which results in their low accuracy.

METHODS

In this paper, we propose a novel method, called MultiSourcDSim, to calculate disease similarity by integrating multiple data sources, namely, gene-disease associations, GO biological process-disease associations and symptom-disease associations. Firstly, we establish three disease similarity networks according to the three disease-related data sources respectively. Secondly, the representation of each node is obtained by integrating the three small disease similarity networks. In the end, the learned representations are applied to calculate the similarity between diseases.

RESULTS

Our approach shows the best performance compared to the other three popular methods. Besides, the similarity network built by MultiSourcDSim suggests that our method can also uncover the latent relationships between diseases.

CONCLUSIONS

MultiSourcDSim is an efficient approach to predict similarity between diseases.

To help aging populations, classify organismal senescence

Comprehensive disease classification and staging is required to address unmet needs of aging populations

Genetic Variation in Human Gene Regulatory Factors Uncovers Regulatory Roles in Local Adaptation and Disease

Differences in gene regulation have been suggested to play essential roles in the evolution of phenotypic changes. Although DNA changes in cis-regulatory elements affect only the regulation of its corresponding gene, variations in gene regulatory factors (trans) can have a broader effect, because the expression of many target genes might be affected. Aiming to better understand how natural selection may have shaped the diversity of gene regulatory factors in human, we assembled a catalog of all proteins involved in controlling gene expression. We found that at least five DNA-binding transcription factor classes are enriched among genes located in candidate regions for selection, suggesting that they might be relevant for understanding regulatory mechanisms involved in human local adaptation. The class of KRAB-ZNFs, zinc-finger (ZNF) genes with a Krüppel-associated box, stands out by first, having the most genes located on candidate regions for positive selection. Second, displaying most nonsynonymous single nucleotide polymorphisms (SNPs) with high genetic differentiation between populations within these regions. Third, having 27 KRAB-ZNF gene clusters with high extended haplotype homozygosity. Our further characterization of nonsynonymous SNPs in ZNF genes located within candidate regions for selection, suggests regulatory modifications that might influence the expression of target genes at population level. Our detailed investigation of three candidate regions revealed possible explanations for how SNPs may influence the prevalence of schizophrenia, eye development, and fertility in humans, among other phenotypes. The genetic variation we characterized here may be responsible for subtle to rough regulatory changes that could be important for understanding human adaptation.

Diagnosis: From classification to prediction

Over the last two decades diagnostic labels have increasingly been sub-divided based on molecular and genetic 'signatures'. But this emphasis on disease sub-types defined in molecular terms, elides the central role of population-based predictive technologies in determining these new diagnoses. While molecular diagnostic sub-types might flow from the laboratory, the clinical validity of every putative diagnostic category must ultimately be tested against its predictive powers. In effect, the former logic of prognosis following diagnosis is reversed. This paper explores the emergence of this new method of diagnostic practice over the last half century.

Automatic classification of free-text medical causes from death certificates for reactive mortality surveillance in France

BACKGROUND

Mortality surveillance is of fundamental importance to public health surveillance. The real-time recording of death certificates, thanks to Electronic Death Registration System (EDRS), provides valuable data for reactive mortality surveillance based on medical causes of death in free-text format. Reactive mortality surveillance is based on the monitoring of mortality syndromic groups (MSGs). An MSG is a cluster of medical causes of death (pathologies, syndromes or symptoms) that meets the objectives of early detection and impact assessment of public health events. The aim of this study is to implement and measure the performance of a rule-based method and two supervised models for automatic free-text cause of death classification from death certificates in order to implement them for routine surveillance.

METHOD

A rule-based method was implemented using four processing steps: standardization rules, splitting causes of death using delimiters, spelling corrections and dictionary projection. A supervised machine learning method using a linear Support Vector Machine (SVM) classifier was also implemented. Two models were produced using different features (SVM1 based solely on surface features and SVM2 combining surface features and MSGs classified by the rule-based method as feature vectors). The evaluation was conducted using an annotated subset of electronic death certificates received between 2012 and 2016. Classification performance was evaluated on seven MSGs (Influenza, Low respiratory diseases, Asphyxia/abnormal respiration, Acute respiratory disease, Sepsis, Chronic digestive diseases, and Chronic endocrine diseases).

RESULTS

The rule-based method and the SVM2 model displayed a high performance with F-measures over 0.94 for all MSGs. Precision and recall were slightly higher for the rule-based method and the SVM2 model. An error-analysis shows that errors were not specific to an MSG.

CONCLUSION

The high performance of the rule-based method and SVM2 model will allow us to set-up a reactive mortality surveillance system based on free-text death certificates. This surveillance will be an added-value for public health decision making.

Analysis of disease organ as a novel phenotype towards disease genetics understanding

Discerning the modular nature of human diseases through computational approaches calls for diverse data. The finding sites of diseases, like other disease phenotypes, possess rich information in understanding disease genetics. Yet, analysis of the rich knowledge of disease finding sites has not been comprehensively investigated. In this study, we built a large-scale disease organ network (DON) based on 76,561 disease-organ associations (for 37,615 diseases and 3492 organs) extracted from the United Medical Language System (UMLS) Metathesaurus. We investigated how phenotypic organ similarity among diseases in DON reflects disease gene sharing. We constructed a disease genetic network (DGN) using curated disease-gene associations and demonstrated that disease pairs with higher organ similarities not only are more likely to share genes, but also tend to share more genes. Based on community detection algorithm, we showed that phenotypic disease clusters on DON significantly correlated with genetic disease clusters on DGN. We compared DON with a state-of-art disease phenotype network, disease manifestation network (DMN), that we have recently constructed, and demonstrated that DON contains complementary knowledge for disease genetics understanding.

The multiplex network of human diseases

Untangling the complex interplay between phenotype and genotype is crucial to the effective characterization and subtyping of diseases. Here we build and analyze the multiplex network of 779 human diseases, which consists of a genotype-based layer and a phenotype-based layer. We show that diseases with common genetic constituents tend to share symptoms, and uncover how phenotype information helps boost genotype information. Moreover, we offer a flexible classification of diseases that considers their molecular underpinnings alongside their clinical manifestations. We detect cohesive groups of diseases that have high intra-group similarity at both the molecular and the phenotypic level. Inspecting these disease communities, we demonstrate the underlying pathways that connect diseases mechanistically. We observe monogenic disorders grouped together with complex diseases for which they increase the risk factor. We propose potentially new disease associations that arise as a unique feature of the information flow within and across the two layers.

Current and Potential Approaches for Defining Disease Signatures: a Systematic Review

Identifying disease signatures in order to facilitate accurate diagnosis/treatment has been the focus of research efforts in the last decade. However, the term "disease signature" has not been properly defined, resulting in inconsistencies between studies, as well as limited ability to fully utilize the tools/information available in the evolving field of healthcare big data. Research was conducted according to the Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA) guidelines. The search (in PubMed, Cochrane, and Web of Science) was limited to English articles published up to 31/12/2016. The search string was "disease signature" OR "disease signatures" OR "disease fingerprint" OR "disease fingerprints" OR "subtype signature" OR "subtype signatures" OR "subgroup signature" OR "subgroup signatures." The full text of the articles was reviewed to determine the meaning of the phrase "disease signature" as well as the context of its use. Of 285 articles identified in the search, 129 were included in the final analysis. The term disease signature was first found in an article from 2001. In the last 10 years, the use of the term increased by approximately ninefold, which is double the general increase in the number of published articles. Only one article attempted to define the term. The two major medical fields where the term was used were oncology (31%) and neurology (20%); 71% of the identified articles used a single biomarker to define the term, 13% of the articles used a pair of biomarkers, and 16% used signatures with multiple biomarker; in 42% of the identified articles, genomic biomarkers were used for the signature, in 17% measurements of biochemical compounds in body fluids, and in 10%, changes in imaging studies were used for the signature. Our findings identified a lack of consistency in defining the term disease signature. We suggest a novel hierarchical multidimensional concept for this term that would combine both current approaches for identifying diseases (one focusing on undesired effects of the disease and the other on its causes). This model can improve disease signature definition consistency which will enable to generalize and classify diseases, resulting in more precise treatments and better outcomes. Ultimately, this model could lead to developing a statistical confidence in a disease signature that would allow physicians/patients to estimate the precision of the diagnosis, which, in turn, may have important implications on patients' prognosis and treatment.

Harnessing formal concepts of biological mechanism to analyze human disease

Mechanism is a widely used concept in biology. In 2017, more than 10% of PubMed abstracts used the term. Therefore, searching for and reasoning about mechanisms is fundamental to much of biomedical research, but until now there has been almost no computational infrastructure for this purpose. Recent work in the philosophy of science has explored the central role that the search for mechanistic accounts of biological phenomena plays in biomedical research, providing a conceptual basis for representing and analyzing biological mechanism. The foundational categories for components of mechanisms—entities and activities—guide the development of general, abstract types of biological mechanism parts. Building on that analysis, we have developed a formal framework for describing and representing biological mechanism, MecCog, and applied it to describing mechanisms underlying human genetic disease. Mechanisms are depicted using a graphical notation. Key features are assignment of mechanism components to stages of biological organization and classes; visual representation of uncertainty, ignorance, and ambiguity; and tight integration with literature sources. The MecCog framework facilitates analysis of many aspects of disease mechanism, including the prioritization of future experiments, probing of gene−drug and gene−environment interactions, identification of possible new drug targets, personalized drug choice, analysis of nonlinear interactions between relevant genetic loci, and classification of diseases based on mechanism.

Naming and the Public Health Roles of Physicians

Resources from the American Medical Association (AMA) Archives facilitate historical consideration of how physicians' authority has been exercised in naming diseases, epidemics, and other health-related issues of national importance. Selected images emphasize physicians' roles in motivating public health initiatives through public service posters, advertisements, and minutes of the AMA House of Delegates meetings.

Opportunities and challenges in developing deep learning models using electronic health records data: a systematic review

Objective

To conduct a systematic review of deep learning models for electronic health record (EHR) data, and illustrate various deep learning architectures for analyzing different data sources and their target applications. We also highlight ongoing research and identify open challenges in building deep learning models of EHRs.

Design/method

We searched PubMed and Google Scholar for papers on deep learning studies using EHR data published between January 1, 2010, and January 31, 2018. We summarize them according to these axes: types of analytics tasks, types of deep learning model architectures, special challenges arising from health data and tasks and their potential solutions, as well as evaluation strategies.

Results

We surveyed and analyzed multiple aspects of the 98 articles we found and identified the following analytics tasks: disease detection/classification, sequential prediction of clinical events, concept embedding, data augmentation, and EHR data privacy. We then studied how deep architectures were applied to these tasks. We also discussed some special challenges arising from modeling EHR data and reviewed a few popular approaches. Finally, we summarized how performance evaluations were conducted for each task.

Discussion

Despite the early success in using deep learning for health analytics applications, there still exist a number of issues to be addressed. We discuss them in detail including data and label availability, the interpretability and transparency of the model, and ease of deployment.

Gene co-expression analysis for functional classification and gene-disease predictions

Gene co-expression networks can be used to associate genes of unknown function with biological processes, to prioritize candidate disease genes or to discern transcriptional regulatory programmes. With recent advances in transcriptomics and next-generation sequencing, co-expression networks constructed from RNA sequencing data also enable the inference of functions and disease associations for non-coding genes and splice variants. Although gene co-expression networks typically do not provide information about causality, emerging methods for differential co-expression analysis are enabling the identification of regulatory genes underlying various phenotypes. Here, we introduce and guide researchers through a (differential) co-expression analysis. We provide an overview of methods and tools used to create and analyse co-expression networks constructed from gene expression data, and we explain how these can be used to identify genes with a regulatory role in disease. Furthermore, we discuss the integration of other data types with co-expression networks and offer future perspectives of co-expression analysis.

Calibration of medical diagnostic classifier scores to the probability of disease

Scores produced by statistical classifiers in many clinical decision support systems and other medical diagnostic devices are generally on an arbitrary scale, so the clinical meaning of these scores is unclear. Calibration of classifier scores to a meaningful scale such as the probability of disease is potentially useful when such scores are used by a physician. In this work, we investigated three methods (parametric, semi-parametric, and non-parametric) for calibrating classifier scores to the probability of disease scale and developed uncertainty estimation techniques for these methods. We showed that classifier scores on arbitrary scales can be calibrated to the probability of disease scale without affecting their discrimination performance. With a finite dataset to train the calibration function, it is important to accompany the probability estimate with its confidence interval. Our simulations indicate that, when a dataset used for finding the transformation for calibration is also used for estimating the performance of calibration, the resubstitution bias exists for a performance metric involving the truth states in evaluating the calibration performance. However, the bias is small for the parametric and semi-parametric methods when the sample size is moderate to large (>100 per class).

Management of minor ailments in a community pharmacy setting: Findings from simulated visits and qualitative study in Gondar town, Ethiopia

Community pharmacy professionals are being widely accepted as sources of treatment and advice for managing minor ailments, largely owing to their location at the heart of the community. The aim of the present study was, therefore, to document the involvement of community pharmacy professionals in the management of minor ailments and perceived barriers that limit their provision of such services. Simulated patient (SP) visits combined with a qualitative study using in-depth interviews was conducted among community pharmacy professionals in Gondar town, Northwest Ethiopia. Scenarios of three different minor ailments (uncomplicated upper respiratory tract infection, back pain and acute diarrhea) were selected and results were reported as percentages. Pharmacy professionals were also interviewed about the barriers in the management of minor ailments. Out of 66 simulated visits, 61 cases (92.4%) provided one or more medications to the SPs. Pharmacy professionals in 16 visits asked SPs information on details of symptoms and past medical and medication history. Ibuprofen alone or in combination with paracetamol was the most commonly dispensed analgesics for back pain. Oral rehydration fluid (ORS) with zinc was the most frequently dispensed medication (33.3%) for the management of acute diarrhea followed by mebendazole (23.9%). Moreover, amoxicillin-clavulanic acid capsule (35%) followed by Amoxicillin (25%) were the most commonly dispensed antibiotics for uncomplicated upper respiratory tract infection. Lack of clinical training and poor community awareness towards the role of community pharmacists in the management of minor ailments were the main barriers for the provision of minor ailment management by community pharmacy professionals. Overall, community pharmacists provided inadequate therapy for the simulated minor ailments. Lack of access to clinical training and poor community awareness were the most commonly cited barriers for providing such services. So as to improve community pharmacists' involvement in managing minor ailments and optimize the contribution of pharmacists, interventions should focus on overcoming the identified barriers.

Designing a more efficient, effective and safe Medical Emergency Team (MET) service using data analysis

Introduction

Hospitals have seen a rise in Medical Emergency Team (MET) reviews. We hypothesised that the commonest MET calls result in similar treatments. Our aim was to design a pre-emptive management algorithm that allowed direct institution of treatment to patients without having to wait for attendance of the MET team and to model its potential impact on MET call incidence and patient outcomes.

Methods

Data was extracted for all MET calls from the hospital database. Association rule data mining techniques were used to identify the most common combinations of MET call causes, outcomes and therapies.

Results

There were 13,656 MET calls during the 34-month study period in 7936 patients. The most common MET call was for hypotension [31%, (2459/7936)]. These MET calls were strongly associated with the immediate administration of intra-venous fluid (70% [1714/2459] v 13% [739/5477] p<0.001), unless the patient was located on a respiratory ward (adjusted OR 0.41 [95%CI 0.25–0.67] p<0.001), had a cardiac cause for admission (adjusted OR 0.61 [95%CI 0.50–0.75] p<0.001) or was under the care of the heart failure team (adjusted OR 0.29 [95%CI 0.19–0.42] p<0.001).

Modelling the effect of a pre-emptive management algorithm for immediate fluid administration without MET activation on data from a test period of 24 months following the study period, suggested it would lead to a 68.7% (2541/3697) reduction in MET calls for hypotension and a 19.6% (2541/12938) reduction in total METs without adverse effects on patients.

Conclusion

Routinely collected data and analytic techniques can be used to develop a pre-emptive management algorithm to administer intravenous fluid therapy to a specific group of hypotensive patients without the need to initiate a MET call. This could both lead to earlier treatment for the patient and less total MET calls.

Social determinants of adult mortality from non-communicable diseases in northern Ethiopia, 2009-2015: Evidence from health and demographic surveillance site

INTRODUCTION

In developing countries, mortality and disability from non-communicable diseases (NCDs) is rising considerably. The effect of social determinants of NCDs-attributed mortality, from the context of developing countries, is poorly understood. This study examines the burden and socio-economic determinants of adult mortality attributed to NCDs in eastern Tigray, Ethiopia.

METHODS

We followed 45,982 adults implementing a community based dynamic cohort design recording mortality events from September 2009 to April 2015. A physician review based Verbal autopsy was used to identify the most probable causes of death. Multivariable Cox proportional hazards regression was performed to identify social determinants of NCD mortality.

RESULTS

Across the 193,758.7 person-years, we recorded 1,091 adult deaths. Compared to communicable diseases, NCDs accounted for a slightly higher proportion of adult deaths; 33% vs 34.5% respectively. The incidence density rate (IDR) of NCD attributed mortality was 194.1 deaths (IDR = 194.1; 95% CI = 175.4, 214.7) per 100,000 person-years. One hundred fifty-seven (41.8%), 68 (18.1%) and 34 (9%) of the 376 NCD deaths were due to cardiovascular disease, cancer and renal failure, respectively. In the multivariable analysis, age per 5-year increase (HR = 1.35; 95% CI: 1.30, 1.41), and extended family and non-family household members (HR = 2.86; 95% CI: 2.05, 3.98) compared to household heads were associated with a significantly increased hazard of NCD mortality. Although the difference was not statistically significant, compared to poor adults, those who were wealthy had a 15% (HR = 0.85; 95% CI: 0.65, 1.11) lower hazard of mortality from NCDs. On the other hand, literate adults (HR = 0.35; 95% CI: 0.13, 0.9) had a significantly decreased hazard of NCD attributed mortality compared to those adults who were unable to read and write. The effect of literacy was modified by age and its effect reduced by 18% for every 5-year increase of age among literate adults.

CONCLUSION

In summary, the study indicates that double mortality burden from both NCDs and communicable diseases was evident in northern rural Ethiopia. Public health intervention measures that prioritise disadvantaged NCD patients such as those who are unable to read and write, the elders, the extended family and non-family household co-residents could significantly reduce NCD mortality among the adult population.

A protein network descriptor server and its use in studying protein, disease, metabolic and drug targeted networks

The genetic, proteomic, disease and pharmacological studies have generated rich data in protein interaction, disease regulation and drug activities useful for systems-level study of the biological, disease and drug therapeutic processes. These studies are facilitated by the established and the emerging computational methods. More recently, the network descriptors developed in other disciplines have become more increasingly used for studying the protein-protein, gene regulation, metabolic, disease networks. There is an inadequate coverage of these useful network features in the public web servers. We therefore introduced upto 313 literature-reported network descriptors in PROFEAT web server, for describing the topological, connectivity and complexity characteristics of undirected unweighted (uniform binding constants and molecular levels), undirected edge-weighted (varying binding constants), undirected node-weighted (varying molecular levels), undirected edge-node-weighted (varying binding constants and molecular levels) and directed unweighted (oriented process) networks. The usefulness of the PROFEAT computed network descriptors is illustrated by their literature-reported applications in studying the protein-protein, gene regulatory, gene co-expression, protein-drug and metabolic networks. PROFEAT is accessible free of charge at http://bidd2.nus.edu.sg/cgi-bin/profeat2016/main.cgi.

What is a disease?

[Full article available at http://rimed.org/rimedicaljournal-2017-10.asp].

Classification of common human diseases derived from shared genetic and environmental determinants

In this study, we used insurance claims for over one-third of the entire US population to create a subset of 128,989 families (481,657 unique individuals). We then used these data to (i) estimate the heritability and familial environmental patterns of 149 diseases and (ii) infer the genetic and environmental correlations for disease pairs from a set of 29 complex diseases. The majority (52 of 65) of our study's heritability estimates matched earlier reports, and 84 of our estimates appear to have been obtained for the first time. We used correlation matrices to compute environmental and genetic disease classifications and corresponding reliability measures. Among unexpected observations, we found that migraine, typically classified as a disease of the central nervous system, appeared to be most genetically similar to irritable bowel syndrome and most environmentally similar to cystitis and urethritis, all of which are inflammatory diseases.

Detecting Spatial Patterns of Disease in Large Collections of Electronic Medical Records Using Neighbor-Based Bootstrapping

We introduce a method called neighbor-based bootstrapping (NB2) that can be used to quantify the geospatial variation of a variable. We applied this method to an analysis of the incidence rates of disease from electronic medical record data (International Classification of Diseases, Ninth Revision codes) for ∼100 million individuals in the United States over a period of 8 years. We considered the incidence rate of disease in each county and its geospatially contiguous neighbors and rank ordered diseases in terms of their degree of geospatial variation as quantified by the NB2 method. We show that this method yields results in good agreement with established methods for detecting spatial autocorrelation (Moran's I method and kriging). Moreover, the NB2 method can be tuned to identify both large area and small area geospatial variations. This method also applies more generally in any parameter space that can be partitioned to consist of regions and their neighbors.