Semantic Changepoint Detection for Finding Potentially Novel Research Publications.

How has the focus of research papers on a given disease changed over time? Identifying the papers at the cusps of change can help highlight the emergence of a new topic or a change in the direction of research. We present a generally applicable unsupervised approach to this question based on semantic changepoints within a given collection of research papers. We illustrate the approach by a range of examples based on a nascent corpus of literature on COVID-19 as well as subsets of papers from PubMed on the World Health Organization list of neglected tropical diseases. The software is freely available at: https://github.com/pdddinakar/SemanticChangepointDetection.

Exposure to NO2, CO, and PM2.5 is linked to regional DNA methylation differences in asthma.

Background: DNA methylation of CpG sites on genetic loci has been linked to increased risk of asthma in children exposed to elevated ambient air pollutants (AAPs). Further identification of specific CpG sites and the pollutants that are associated with methylation of these CpG sites in immune cells could impact our understanding of asthma pathophysiology. In this study, we sought to identify some CpG sites in specific genes that could be associated with asthma regulation (Foxp3 and IL10) and to identify the different AAPs for which exposure prior to the blood draw is linked to methylation levels at these sites. We recruited subjects from Fresno, California, an area known for high levels of AAPs. Blood samples and responses to questionnaires were obtained (n = 188), and in a subset of subjects (n = 33), repeat samples were collected 2 years later. Average measures of AAPs were obtained for 1, 15, 30, 90, 180, and 365 days prior to each blood draw to estimate the short-term vs. long-term effects of the AAP exposures. Results: Asthma was significantly associated with higher differentially methylated regions (DMRs) of the Foxp3 promoter region (p = 0.030) and the IL10 intronic region (p = 0.026). Additionally, at the 90-day time period (90 days prior to the blood draw), Foxp3 methylation was positively associated with NO2, CO, and PM2.5 exposures (p = 0.001, p = 0.001, and p = 0.012, respectively). In the subset of subjects retested 2 years later (n = 33), a positive association between AAP exposure and methylation was sustained. There was also a negative correlation between the average Foxp3 methylation of the promoter region and activated Treg levels (p = 0.039) and a positive correlation between the average IL10 methylation of region 3 of intron 4 and IL10 cytokine expression (p = 0.030). Conclusions: Short-term and long-term exposures to high levels of CO, NO2, and PM2.5 were associated with alterations in differentially methylated regions of Foxp3. IL10 methylation showed a similar trend. For any given individual, these changes tend to be sustained over time. In addition, asthma was associated with higher differentially methylated regions of Foxp3 and IL10.

Erratum: Constellation: a tool for rapid, automated phenotype assignment of a highly polymorphic pharmacogene, CYP2D6, from whole-genome sequences.

[This corrects the article DOI: 10.1038/npjgenmed.2015.7.].

Automated Prediction of Hepatic Arterial Stenosis.

Several thousand life-saving liver transplants are performed each year. One of the most common causes of early transplant failure is arterial stenosis of the anastomotic junction. Early detection of transplant arterial stenosis can help prevent transplant failure and the need to re-transplant. Doppler ultrasound is the most common screening method, but it suffers from poor specificity. Positive screening cases proceed to angiography which is an invasive and expensive procedure. A more accurate test could decrease the number of normal patients who would have to undergo this invasive diagnostic procedure. We present a turnkey clinical decision support tool for automated prediction of stenosis based on Fourier spectrum analysis of Doppler sonograms to compute a Stenosis Index that has been shown to have higher accuracy than traditional measures. The results of the automated approach compare favorably with the manual approach. Software is available from the authors on request.

Engineered Tissue Inhibitor of Metalloproteinases-3 Variants Resistant to Endocytosis Have Prolonged Chondroprotective Activity.

Tissue inhibitor of metalloproteinases-3 (TIMP-3) is a central inhibitor of matrix-degrading and sheddase families of metalloproteinases. Extracellular levels of the inhibitor are regulated by the balance between its retention on the extracellular matrix and its endocytic clearance by the scavenger receptor low density lipoprotein receptor-related protein 1 (LRP1). Here, we used molecular modeling to predict TIMP-3 residues potentially involved in binding to LRP1 based on the proposed LRP1 binding motif of 2 lysine residues separated by about 21 Å and mutated the candidate lysine residues to alanine individually and in pairs. Of the 22 mutants generated, 13 displayed a reduced rate of uptake by HTB94 chondrosarcoma cells. The two mutants (TIMP-3 K26A/K45A and K42A/K110A) with lowest rates of uptake were further evaluated and found to display reduced binding to LRP1 and unaltered inhibitory activity against prototypic metalloproteinases. TIMP-3 K26A/K45A retained higher affinity for sulfated glycosaminoglycans than K42A/K110A and exhibited increased affinity for ADAMTS-5 in the presence of heparin. Both mutants inhibited metalloproteinase-mediated degradation of cartilage at lower concentrations and for longer than wild-type TIMP-3, indicating that their increased half-lives improved their ability to protect cartilage. These mutants may be useful in treating connective tissue diseases associated with increased metalloproteinase activity.

Association of tree nut and coconut sensitizations.

BACKGROUND: Coconut (Cocos nucifera), despite being a drupe, was added to the US Food and Drug Administration list of tree nuts in 2006, causing potential confusion regarding the prevalence of coconut allergy among tree nut allergic patients. OBJECTIVE: To determine whether sensitization to tree nuts is associated with increased odds of coconut sensitization. METHODS: A single-center retrospective analysis of serum specific IgE levels to coconut, tree nuts (almond, Brazil nut, cashew, chestnut, hazelnut, macadamia, pecan, pistachio, and walnut), and controls (milk and peanut) was performed using deidentified data from January 2000 to August 2012. Spearman correlation (ρ) between coconut and each tree nut was determined, followed by hierarchical clustering. Sensitization was defined as a nut specific IgE level of 0.35 kU/L or higher. Unadjusted and adjusted associations between coconut and tree nut sensitization were tested by logistic regression. RESULTS: Of 298 coconut IgE values, 90 (30%) were considered positive results, with a mean (SD) of 1.70 (8.28) kU/L. Macadamia had the strongest correlation (ρ = 0.77), whereas most other tree nuts had significant (P < .05) but low correlation (ρ < 0.5) with coconut. The adjusted odds ratio between coconut and macadamia was 7.39 (95% confidence interval, 2.60-21.02; P < .001) and 5.32 (95% confidence interval, 2.18-12.95; P < .001) between coconut and almond, with other nuts not being statistically significant. CONCLUSION: Our findings suggest that although sensitization to most tree nuts appears to correlate with coconut, this is largely explained by sensitization to almond and macadamia. This finding has not previously been reported in the literature. Further study correlating these results with clinical symptoms is planned.

Constellation: a tool for rapid, automated phenotype assignment of a highly polymorphic pharmacogene, CYP2D6, from whole-genome sequences.

An important component of precision medicine-the use of whole-genome sequencing (WGS) to guide lifelong healthcare-is electronic decision support to inform drug choice and dosing. To achieve this, automated identification of genetic variation in genes involved in drug absorption, distribution, metabolism, excretion and response (ADMER) is required. CYP2D6 is a major enzyme for drug bioactivation and elimination. CYP2D6 activity is predominantly governed by genetic variation; however, it is technically arduous to haplotype. Not only is the nucleotide sequence of CYP2D6 highly polymorphic, but the locus also features diverse structural variations, including gene deletion, duplication, multiplication events and rearrangements with the nonfunctional, neighbouring CYP2D7 and CYP2D8 genes. We developed Constellation, a probabilistic scoring system, enabling automated ascertainment of CYP2D6 activity scores from 2×100 paired-end WGS. The consensus reference method included TaqMan genotyping assays, quantitative copy-number variation determination and Sanger sequencing. When compared with the consensus reference Constellation had an analytic sensitivity of 97% (59 of 61 diplotypes) and analytic specificity of 95% (116 of 122 haplotypes). All extreme phenotypes, i.e., poor and ultrarapid metabolisers were accurately identified by Constellation. Constellation is anticipated to be extensible to functional variation in all ADMER genes, and to be performed at marginal incremental financial and computational costs in the setting of diagnostic WGS.

Finding disease similarity based on implicit semantic similarity.

Genomics has contributed to a growing collection of gene-function and gene-disease annotations that can be exploited by informatics to study similarity between diseases. This can yield insight into disease etiology, reveal common pathophysiology and/or suggest treatment that can be appropriated from one disease to another. Estimating disease similarity solely on the basis of shared genes can be misleading as variable combinations of genes may be associated with similar diseases, especially for complex diseases. This deficiency can be potentially overcome by looking for common biological processes rather than only explicit gene matches between diseases. The use of semantic similarity between biological processes to estimate disease similarity could enhance the identification and characterization of disease similarity. We present functions to measure similarity between terms in an ontology, and between entities annotated with terms drawn from the ontology, based on both co-occurrence and information content. The similarity measure is shown to outperform other measures used to detect similarity. A manually curated dataset with known disease similarities was used as a benchmark to compare the estimation of disease similarity based on gene-based and Gene Ontology (GO) process-based comparisons. The detection of disease similarity based on semantic similarity between GO Processes (Recall=55%, Precision=60%) performed better than using exact matches between GO Processes (Recall=29%, Precision=58%) or gene overlap (Recall=88% and Precision=16%). The GO-Process based disease similarity scores on an external test set show statistically significant Pearson correlation (0.73) with numeric scores provided by medical residents. GO-Processes associated with similar diseases were found to be significantly regulated in gene expression microarray datasets of related diseases.

Drug repositioning using disease associated biological processes and network analysis of drug targets.

The analysis of disease using protein-protein interaction networks and network pharmacology has enabled better understanding of disease etiology and drug action. New insights into disease etiology and a better understanding of biological subsystems have opened up the possibility of finding new uses for existing drugs besides their original medical indication. We present an approach which makes use of the biological processes associated with diseases along with their known drugs and drug targets to predict Biological Process-Drug relationships. Network analysis is used to further refine these associations to eventually predict new Disease-Drug relationships. The approach is validated by the observation that, out of 2078 predicted disease-drug relationships, 401 (18.1%) have been used in a clinical trial.

Automated ontological gene annotation for computing disease similarity.

The annotation of gene/gene products with information on associated diseases is useful as an aid to clinical diagnosis and drug discovery. Several supervised and unsupervised methods exist that automate the association of genes with diseases, but relatively little work has been done to map protein sequence data to disease terminologies. This paper augments an existing open-disease terminology, the Disease Ontology (DO), and uses it for automated annotation of Swissprot records. In addition to the inherent benefits of mapping data to a rich ontology, we demonstrate a gain of 36.1% in gene-disease associations compared to that in DO. Further, we measure disease similarity by exploiting the co-occurrence of annotation among proteins and the hierarchical structure of DO. This makes it possible to find related diseases or signs, with the potential to find previously unknown relationships.

MindTrial: An Intelligent System for Clinical Trials.

The recruitment of human subjects for clinical trials research is a critically important step in the discovery of new cures for diseases. However, the current recruitment methodologies are inherently inefficient. Considerable resources are expended in efforts to recruit adequate numbers of patient volunteers who meet the inclusion/exclusion criteria for clinical trials. Recruitment is particularly challenging for trials involving vulnerable, psychiatrically disordered groups. We have developed a prototype system, called MindTrial, that is based on an online model to enhance the efficiency and quality of recruitment of patients with psychiatric disorders for clinical research. The intelligent component of the MindTrial system can facilitate highly specific matches between clinical trial criteria and volunteers for self-enrollment of sufficient numbers of patient volunteers. We believe this system is particularly valuable in optimizing recruitment for clinical trial studies for development of new drugs.

A pragmatic approach to mapping the open biomedical ontologies.

There has been a large increase in the number of ontologies that have been introduced by the biomedical community in recent years. To maximise their potential, there is an urgent need for a mechanism to build interoperability between ontologies developed by different groups. While identifying and linking related concepts is of obvious importance, it is also essential to analyse how ontologies as a whole overlap and can be clustered. This paper explores overlapping relationships in the Open Biomedical Ontologies (OBO) and provides an interoperability framework called InterOBO for sharing biomedical knowledge across OBO communities.

Applications of medical informatics in allergy/immunology.

OBJECTIVE: To provide a general overview of informatics and its interface with allergy/immunology. DATA SOURCES: The PubMed interface to MEDLINE was searched with the keywords asthma, allergy, or immunology together with the keywords informatics, bioinformatics, and information technology to retrieve the articles relevant to this review. STUDY SELECTION: The authors' knowledge of the field was used to include sources of information other than those obtained through the MEDLINE search. RESULTS: A survey of informatics, with emphasis on the relevance to allergy, asthma, and immunology, is presented. CONCLUSIONS: Several innovative informatics approaches have significant potential to improve health care on diverse fronts. Newer methods of information representation are poised to facilitate the impact of cutting-edge research on clinical practice.

MachineProse: an ontological framework for scientific assertions.

OBJECTIVE: The idea of testing a hypothesis is central to the practice of biomedical research. However, the results of testing a hypothesis are published mainly in the form of prose articles. Encoding the results as scientific assertions that are both human and machine readable would greatly enhance the synergistic growth and dissemination of knowledge. DESIGN: We have developed MachineProse (MP), an ontological framework for the concise specification of scientific assertions. MP is based on the idea of an assertion constituting a fundamental unit of knowledge. This is in contrast to current approaches that use discrete concept terms from domain ontologies for annotation and assertions are only inferred heuristically. MEASUREMENTS: We use illustrative examples to highlight the advantages of MP over the use of the Medical Subject Headings (MeSH) system and keywords in indexing scientific articles. RESULTS: We show how MP makes it possible to carry out semantic annotation of publications that is machine readable and allows for precise search capabilities. In addition, when used by itself, MP serves as a knowledge repository for emerging discoveries. A prototype for proof of concept has been developed that demonstrates the feasibility and novel benefits of MP. As part of the MP framework, we have created an ontology of relationship types with about 100 terms optimized for the representation of scientific assertions. CONCLUSION: MachineProse is a novel semantic framework that we believe may be used to summarize research findings, annotate biomedical publications, and support sophisticated searches.

Ontological modeling of transformation in heart defect diagrams.

The accurate portrayal of a large volume data of variable heart defects is crucial to providing good patient care in pediatric cardiology. Our research aims to span the universe of congenital heart defects by generating illustrative diagrams that enhance data interpretation. To accommodate the range and severity of defects to be represented, we base our diagrams on transformation models applied to a normal heart rather than a static set of defects. These models are based on a domain-specific ontology, clustering, association rule mining and the use of parametric equations specified in a mathematical programming language.

Tandem machine learning for the identification of genes regulated by transcription factors.

BACKGROUND: The identification of promoter regions that are regulated by a given transcription factor has traditionally relied upon the identification and distributions of binding sites recognized by the factor. In this study, we have developed a tandem machine learning approach for the identification of regulatory target genes based on these parameters and on the corresponding binding site information contents that measure the affinities of the factor for these cognate elements. RESULTS: This method has been validated using models of DNA binding sites recognized by the xenobiotic-sensitive nuclear receptor, PXR/RXRalpha, for target genes within the human genome. An information theory-based weight matrix was first derived and refined from known PXR/RXRalpha binding sites. The promoter region of candidate genes was scanned with the weight matrix. A novel information density-based clustering algorithm was then used to identify clusters of information rich sites. Finally, transformed data representing metrics of location, strength and clustering of binding sites were used for classification of promoter regions using an ensemble approach involving neural networks, decision trees and Naïve Bayesian classification. The method was evaluated on a set of 24 known target genes and 288 genes known not to be regulated by PXR/RXRalpha. We report an average accuracy (proportion of correctly classified promoter regions) of 71%, sensitivity of 73%, and specificity of 70%, based on multiple cross-validation and the leave-one-out strategy. The performance on a test set of 13 genes showed that 10 were correctly classified. CONCLUSION: We have developed a machine learning approach for the successful detection of gene targets for transcription factors with high accuracy. The method has been validated for the transcription factor PXR/RXRalpha and has the potential to be extended to other transcription factors.

An informatics search for the low-molecular weight chromium-binding peptide.

BACKGROUND: The amino acid composition of a low molecular weight chromium binding peptide (LMWCr), isolated from bovine liver, is reportedly E:G:C:D::4:2:2:2, though its sequence has not been discovered. There is some controversy surrounding the exact biochemical forms and the action of Cr(III) in biological systems; the topic has been the subject of many experimental reports and continues to be investigated. Clarification of Cr-protein interactions will further understanding Cr(III) biochemistry and provide a basis for novel therapies based on metallocomplexes or small molecules. RESULTS: A genomic search of the non-redundant database for all possible decapeptides of the reported composition yields three exact matches, EDGEECDCGE, DGEECDCGEE and CEGGCEEDDE. The first two sequences are found in ADAM 19 (A Disintegrin and Metalloproteinase domain 19) proteins in man and mouse; the last is found in a protein kinase in rice (Oryza sativa). A broader search for pentameric sequences (and assuming a disulfide dimer) corresponding to the stoichiometric ratio E:D:G:C::2:1:1:1, within the set of human proteins and the set of proteins in, or related to, the insulin signaling pathway, yields a match at an acidic region in the alpha-subunit of the insulin receptor (-EECGD-, residues 175-184). A synthetic peptide derived from this sequence binds chromium(III) and forms a metal-peptide complex that has properties matching those reported for isolated LMWCr and Cr(III)-containing peptide fractions. CONCLUSION: The search for an acidic decameric sequence indicates that LMWCr may not be a contiguous sequence. The identification of a distinct pentameric sequence in a significant insulin-signaling pathway protein suggests a possible identity for the LMWCr peptide. This identification clarifies directions for further investigation of LMWCr peptide fractions, chromium bio-coordination chemistry and a possible role in the insulin signaling pathway. Implications for models of chromium action in the insulin-signaling pathway are discussed.