A global network of biomedical relationships derived from text

Percha, Bethany; Altman, Russ B.

doi:10.1093/bioinformatics/bty114

Cited by 120 publications

(99 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Pubtator, a tool designed to annotate biological terms cited in PubMed documents, was used to annotate the biomarkers cited within the literature retrieved from the three queries (14). The…”

Section: Biomarker Annotation and Extractionmentioning

confidence: 99%

Overlapping Pathogenic Signalling Pathways and Biomarkers in Preeclampsia and Cardiovascular Disease

Šuvakov

Bonner

Nikolić

et al. 2020

Preprint

View full text Add to dashboard Cite

Corresponding authorsHighlights  Women with preeclampsia have increased risk of cardiovascular disease later in life but the mechanism is poorly understood. "In silico" analyses of publically available datasets provided overlapping biomarkers and pathogenic pathways between preeclampsia, hypertension and heart failure with preserved ejection fraction (HFpEF). These data could be utilised in the future studies that may lead to the development of better risk stratification strategies or preventative treatments for women post preeclampsia. AbstractObjectives: Preeclampsia is a cardiovascular pregnancy complication which occurs in 5-10% of pregnancies that can lead to a number of pregnancy complications including maternal and foetal death. Long-term, preeclampsia is associated with up to 8-fold increased risk of cardiovascular disease (CVD) for both mothers and their offspring. The lack of mechanistic data in relation to the causes or consequences of preeclampsia has prevented the development of effective therapeutic or monitoring strategies. Study design:This study investigates common underlying mechanisms of preeclampsia and CVD, specifically hypertension and heart failure with preserved ejection fraction (HFpEF)using "in silico" approach of publicly available datasets. Integrated techniques were designed to mine data repositories and identify relevant biomarkers associated with these three conditions. Main outcomes measures:The knowledge base tools were employed that enabled the analysis of these biomarkers to discover potential molecular and biological links between these three conditions.Results: Our bioinformatics "in silico" analyses of the publically available datasets identified 76 common biomarkers between preeclampsia, hypertension and HFpEF. These biomarkers were representative of 29 pathways commonly enriched across the three conditions which were largely related to inflammation, metabolism, angiogenesis, remodelling, haemostasis, apoptosis, endoplasmic reticulum (ER) stress signalling and the renin-angiotensin-aldosterone (RAAS) system. Conclusions:This bioinformatics approach which uses the wealth of scientific data available in public repositories can be helpful to gain a deeper understanding of the overlapping pathogenic mechanisms of associated diseases, which could be explored as biomarkers or targets to prevent long-term cardiovascular complications such as hypertension and HFpEF following preeclampsia. Abstract word count: 244 wordsMain text word count: 2580 words Total word count: 4653 words

show abstract

Section: Biomarker Annotation and Extractionmentioning

confidence: 99%

Overlapping Pathogenic Signalling Pathways and Biomarkers in Preeclampsia and Cardiovascular Disease

Šuvakov

Bonner

Nikolić

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…A well known body of work, OpenIE (Banko et al, 2007;Fader et al, 2011;Mausam et al, 2012;Angeli et al, 2015) aims to extract patterns between entity mentions in sentences, thereby discovering new surface forms which can be clustered (Mohamed et al, 2011;Nakashole et al, 2012) in order to reveal new meaningful relationship types. In the biomedical domain, Percha and Altman (2018) attempt something similar by extracting and clustering dependency patterns between pairs of biomedical entities (e.g. chemicalgene, chemical-disease, gene-disease).…”

Section: Related Workmentioning

confidence: 99%

Constructing large scale biomedical knowledge bases from scratch with rapid annotation of interpretable patterns

Fauqueur

Thillaisundaram²,

Togia

2019

Proceedings of the 18th BioNLP Workshop and Shared Task

View full text Add to dashboard Cite

Knowledge base construction is crucial for summarising, understanding and inferring relationships between biomedical entities. However, for many practical applications such as drug discovery, the scarcity of relevant facts (e.g. gene X is therapeutic target for disease Y) severely limits a domain expert's ability to create a usable knowledge base, either directly or by training a relation extraction model. In this paper, we present a simple and effective method of extracting new facts with a pre-specified binary relationship type from the biomedical literature, without requiring any training data or hand-crafted rules. Our system discovers, ranks and presents the most salient patterns to domain experts in an interpretable form. By marking patterns as compatible with the desired relationship type, experts indirectly batch-annotate candidate pairs whose relationship is expressed with such patterns in the literature. Even with a complete absence of seed data, experts are able to discover thousands of high-quality pairs with the desired relationship within minutes. When a small number of relevant pairs do exist -even when their relationship is more general (e.g. gene X is biologically associated with disease Y) than the relationship of interest -our system leverages them in order to i) learn a better ranking of the patterns to be annotated or ii) generate weakly labelled pairs in a fully automated manner. We evaluate our method both intrinsically and via a downstream knowledge * Equal contribution. Listing order is alphabetical. Theodosia proposed and co-ordinated the research project, built the early prototypes and contributed the different methods for extracting and lexicalising patterns. Ashok provided conceptual work on the metrics for ranking simplifications and for the intrinsic evaluation, developed the simplification extraction module, ran the experiments for the automated workflow (with all the parameter variations) and performed all the extrinsic evaluations. Julien was mainly responsible for the system architecture and workflow, the intrinsic evaluation (including interacting with the experts), handling negation and speculation and the clustering algorithm. base completion task, and show that it is an effective way of constructing knowledge bases when few or no relevant facts are already available.

show abstract

“…Clustering is an unsupervised approach that extracts relationships from text by grouping similar sentences together. Percha et al used this technique to group sentences based on their grammatical structure [31]. Using Stanford's Core NLP Parser [32], a dependency tree was generated for every sentence in each Pubmed abstract [31].…”

Section: Unsupervised Extractorsmentioning

confidence: 99%

“…Percha et al used this technique to group sentences based on their grammatical structure [31]. Using Stanford's Core NLP Parser [32], a dependency tree was generated for every sentence in each Pubmed abstract [31]. Each tree was clustered based on similarity and each cluster was manually annotated to determine which relationship each group represented [31].…”

Section: Unsupervised Extractorsmentioning

confidence: 99%

“…Using Stanford's Core NLP Parser [32], a dependency tree was generated for every sentence in each Pubmed abstract [31]. Each tree was clustered based on similarity and each cluster was manually annotated to determine which relationship each group represented [31]. For our project we incorporated the results of this work as domain heuristic label functions.…”

Section: Unsupervised Extractorsmentioning

confidence: 99%

See 1 more Smart Citation

Expanding a Database-derived Biomedical Knowledge Graph via Multi-relation Extraction from Biomedical Abstracts

Nicholson

Himmelstein

Greene

2019

Preprint

View full text Add to dashboard Cite

Knowledge graphs support multiple research efforts by providing contextual information for biomedical entities, constructing networks, and supporting the interpretation of high-throughput analyses. These databases are populated via some form of manual curation, which is difficult to scale in the context of an increasing publication rate. Data programming is a paradigm that circumvents this arduous manual process by combining databases with simple rules and heuristics written as label functions, which are programs designed to automatically annotate textual data. Unfortunately, writing a useful label function requires substantial error analysis and is a nontrivial task that takes multiple days per function. This makes populating a knowledge graph with multiple nodes and edge types practically infeasible. We sought to accelerate the label function creation process by evaluating the extent to which label functions could be re-used across multiple edge types. We used a subset of an existing knowledge graph centered on disease, compound, and gene entities to evaluate label function re-use. We determined the best label function combination by comparing a baseline database-only model with the same model but added edge-specific or edge-mismatch label functions. We confirmed that adding additional edge-specific rather than edge-mismatch label functions often improves text annotation and shows that this approach can incorporate novel edges into our source knowledge graph. We expect that continued development of this strategy has the potential to swiftly populate knowledge graphs with new discoveries, ensuring that these resources include cutting-edge results.

show abstract

A global network of biomedical relationships derived from text

Cited by 120 publications

References 44 publications

Overlapping Pathogenic Signalling Pathways and Biomarkers in Preeclampsia and Cardiovascular Disease

Overlapping Pathogenic Signalling Pathways and Biomarkers in Preeclampsia and Cardiovascular Disease

Constructing large scale biomedical knowledge bases from scratch with rapid annotation of interpretable patterns

Expanding a Database-derived Biomedical Knowledge Graph via Multi-relation Extraction from Biomedical Abstracts

Contact Info

Product

Resources

About