LinkExplorer: Predicting, explaining and exploring links in large biomedical knowledge graphs

Machine learning algorithms for link prediction can be valuable tools for hypothesis generation. However, many current algorithms are black boxes or lack good user interfaces that could facilitate insight into why predictions are made. We present LinkExplorer, a software suite for predicting, explaining and exploring links in large biomedical knowledge graphs. LinkExplorer integrates our novel, rule-based link prediction engine SAFRAN, which was recently shown to outcompete other explainable algorithms and established black box algorithms. Here, we demonstrate highly competitive evaluation results of our algorithm on multiple large biomedical knowledge graphs, and release a web interface that allows for interactive and intuitive exploration of predicted links and their explanations.

A graph-embedded topic model enables characterization of diverse pain phenotypes among UK Biobank individuals

Large biobank repositories of clinical conditions and medications data open opportunities to investigate the phenotypic disease network. To enable systematic investigation of entire structured phenomes, we present graph embedded topic model (GETM). We offer two main methodological contributions in GETM. First, to aid topic inference, we integrate existing biomedical knowledge graph information in the form of pre-trained graph embedding into the embedded topic model. Second, leveraging deep learning techniques, we developed a variational autoencoder framework to infer patient phenotypic mixture. For interpretability, we use a linear decoder to simultaneously infer the bi-modal distributions of the disease conditions and medications. We applied GETM to UK Biobank (UKB) self-reported clinical phenotype data, which contains conditions and medications for 457,461 individuals. Compared to existing methods, GETM demonstrates overall superior performance in imputing missing conditions and medications. Here, we focused on characterizing pain phenotypes recorded in the questionnaire of the UKB individuals. GETM accurately predicts the status of chronic musculoskeletal (CMK) pain, chronic pain by body-site, and non-specific chronic pain using past conditions and medications. Our analyses revealed not only the known pain-related topics but also the surprising predominance of medications and conditions in the cardiovascular category among the most predictive topics across chronic pain phenotypes.

Key concepts for searching evidence: an introduction for healthcare professionals

The currently abundant bibliography on healthcare can make the search process an exhausting and frustrating experience. For this reason, it is essential to learn the basic concepts of research question formulation, information sources, and search strategies to make this process more efficient and user-friendly. The search strategy is an iterative process that allows the incorporation of tools and terms in the strategy design to optimize evidence retrieval. Each strategy varies according to the questions, the language used, the source of information accessed, and the available tools. This article is part of a methodological series of narrative reviews on biostatistics and clinical epidemiology. This narrative review describes the essential elements for developing a literature search strategy and identifying the relevant evidence concerning a clinical question through familiar and accessible sources (such as Google and Google Scholar), as well as search interfaces and technical-scientific databases focused on biomedical knowledge (PubMed and The Cochrane Library).

130 Impact of Low-Dose High-Frequency in situ Simulation on Inpatient Diabetes Management: A Pilot Study

Even in the presence of established institutional guidelines, failure of compliance by the clinical teams plays an important role in the control of diabetes. The identified gaps include contextual and biomedical knowledge, attitudes, clinical inertia, confidence and familiarity with existing hospital resources and guidelines with regards to hospital diabetes care We wanted to demonstrate the efficacy of low-dose high-frequency The exercise was a 15-minute session, delivered during working hours to individual nurses. This consisted of a 5-minute scenario, involving a standardized patient followed by a 10-minute debrief. Modified Diamond-model debrief with an advocacy-inquiry model was used by the debriefer, a trained fellow in simulation, and overseen by an expert. The scripted scenario involved a patient with Diabetic Ketoacidosis (DKA), with learning outcomes of recognizing DKA, managing the patient and adhering to the institutional guidelines including management of hypoglycaemia. The scenario was individualized based on the roles of the participants. Pre- and post-questionnaires were given to the participants. The simulation was repeated twice in the second week and once in the third week.This mixed-method study was conducted in a UK teaching hospital, in a ward designated for patients with diabetes, as a part of a quality improvement programme. In the first week, patients with diabetes, admitted for DKA, were chosen and their blood sugar recordings, dysglycaemic episodes and adherence to guidelines were noted. Every week data were collected as in the first week. GNU pspp 1.0.1 [version 3] free software was used. The confidence scores were given as mean and standard deviation with confidence interval (CI) of 98.75%. A p-value of <0.0125 was considered significant based on the number of data points.The Dysglycemic episodes and protocol adherence from medical recordsConsidering the T2 (increased recognition of diabetic emergencies and adherence to protocol) and T3 (improved patient outcomes) outcomes, the methodology was recommended as a modality of training the nursing staff involved in inpatient care of patients with diabetes. Future programmes including multi-disciplinary teams, to explore teamwork and communication, are planned.

Human studies of mitochondrial biology demonstrate an overall lack of binary sex differences: A multivariate meta-analysis

Mitochondria are maternally inherited organelles that play critical tissue-specific roles, including hormone synthesis and energy production, that influence development, health, and aging. However, whether mitochondria from women and men exhibit consistent biological differences remains unclear, representing a major gap in biomedical knowledge. This meta-analysis systematically examined 4 domains and 6 subdomains of mitochondrial biology (total 39 measures), including mitochondrial content, respiratory capacity, reactive oxygen species (ROS) production, morphometry, and mitochondrial DNA copy number. Standardized effect sizes (Hedges g) of sex differences were computed for each measure using data in 2,258 participants (51.5% women) from 50 studies. Only two measures demonstrated aggregate binary sex differences: higher mitochondrial content in women (g = 0.20, chi2 p = 0.01), and higher ROS production in skeletal muscle in men (g = 0.49, chi2 p < 0.0001). Sex differences showed weak to no correlation with age or BMI. Studies with small sample sizes tended to overestimate effect sizes (r = -0.17, p < 0.001), and sex differences varied by tissue examined. Our findings point to a wide variability of findings in the literature concerning possible binary sex differences in mitochondrial biology. Studies specifically designed to capture sex- and gender-related differences in mitochondrial biology are needed, including detailed considerations of physical activity and sex hormones.

Identification of missing concepts in biomedical terminologies using sequence-based formal concept analysis

Background As biomedical knowledge is rapidly evolving, concept enrichment of biomedical terminologies is an active research area involving automatic identification of missing or new concepts. Previously, we prototyped a lexical-based formal concept analysis (FCA) approach in which concepts were derived by intersecting bags of words, to identify potentially missing concepts in the National Cancer Institute (NCI) Thesaurus. However, this prototype did not handle concept naming and positioning. In this paper, we introduce a sequenced-based FCA approach to identify potentially missing concepts, supporting concept naming and positioning. Methods We consider the concept name sequences as FCA attributes to construct the formal context. The concept-forming process is performed by computing the longest common substrings of concept name sequences. After new concepts are formalized, we further predict their potential positions in the original hierarchy by identifying their supertypes and subtypes from original concepts. Automated validation via external terminologies in the Unified Medical Language System (UMLS) and biomedical literature in PubMed is performed to evaluate the effectiveness of our approach. Results We applied our sequenced-based FCA approach to all the sub-hierarchies under Disease or Disorder in the NCI Thesaurus (19.08d version) and five sub-hierarchies under Clinical Finding and Procedure in the SNOMED CT (US Edition, March 2020 release). In total, 1397 potentially missing concepts were identified in the NCI Thesaurus and 7223 in the SNOMED CT. For NCI Thesaurus, 85 potentially missing concepts were found in external terminologies and 315 of the remaining 1312 appeared in biomedical literature. For SNOMED CT, 576 were found in external terminologies and 1159 out of the remaining 6647 were found in biomedical literature. Conclusion Our sequence-based FCA approach has shown the promise for identifying potentially missing concepts in biomedical terminologies.