A Knowledge Graph of Combined Drug Therapies Using Semantic Predications From Biomedical Literature: Algorithm Development

Background Combination therapy plays an important role in the effective treatment of malignant neoplasms and precision medicine. Numerous clinical studies have been carried out to investigate combination drug therapies. Automated knowledge discovery of these combinations and their graphic representation in knowledge graphs will enable pattern recognition and identification of drug combinations used to treat a specific type of cancer, improve drug efficacy and treatment of human disorders. Objective This paper aims to develop an automated, visual approach to discover knowledge about combination therapies from biomedical literature, especially from those studies with high-level evidence such as clinical trial reports and clinical practice guidelines. Methods Based on semantic predications, which consist of a triple structure of subject-predicate-object (SPO), we proposed an automated algorithm to discover knowledge of combination drug therapies using the following rules: 1) two or more semantic predications (S1-P-O and Si-P-O, i = 2, 3…) can be extracted from one conclusive claim (sentence) in the abstract of a given publication, and 2) these predications have an identical predicate (that closely relates to human disease treatment, eg, “treat”) and object (eg, disease name) but different subjects (eg, drug names). A customized knowledge graph organizes and visualizes these combinations, improving the traditional semantic triples. After automatic filtering of broad concepts such as “pharmacologic actions” and generic disease names, a set of combination drug therapies were identified and characterized through manual interpretation. Results We retrieved 22,263 clinical trial reports and 31 clinical practice guidelines from PubMed abstracts by searching “antineoplastic agents” for drug restriction (published between Jan 2009 and Oct 2019). There were 15,603 conclusive claims locally parsed using the search terms “conclusion*” and “conclude*” ready for semantic predications extraction by SemRep, and 325 candidate groups of semantic predications about combined medications were automatically discovered within 316 conclusive claims. Based on manual analysis, we determined that 255/316 claims (78.46%) were accurately identified as describing combination therapies and adopted these to construct the customized knowledge graph. We also identified two categories (and 4 subcategories) to characterize the inaccurate results: limitations of SemRep and limitations of proposal. We further learned the predominant patterns of drug combinations based on mechanism of action for new combined medication studies and discovered 4 obvious markers (“combin*,” “coadministration,” “co-administered,” and “regimen”) to identify potential combination therapies to enable development of a machine learning algorithm. Conclusions Semantic predications from conclusive claims in the biomedical literature can be used to support automated knowledge discovery and knowledge graph construction for combination therapies. A machine learning approach is warranted to take full advantage of the identified markers and other contextual features.

[1]  Cui Tao,et al.  Constructing Biomedical Knowledge Graph Based on SemMedDB and Linked Open Data , 2018, 2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM).

[2]  Erik M. van Mulligen,et al.  Using predicate and provenance information from a knowledge graph for drug efficacy screening , 2018, Journal of Biomedical Semantics.

[3]  Hua Xu,et al.  Automated identification of molecular effects of drugs (AIMED) , 2016, J. Am. Medical Informatics Assoc..

[4]  Sri Nurdiati,et al.  25 years development of knowledge graph theory: the results and the challenge , 2008 .

[5]  Marcelo Fiszman,et al.  The interaction of domain knowledge and linguistic structure in natural language processing: interpreting hypernymic propositions in biomedical text , 2003, J. Biomed. Informatics.

[6]  Ayesha Salahuddin,et al.  Combination therapy for hypertension 2013: an update. , 2013, Journal of the American Society of Hypertension : JASH.

[7]  João D. Ferreira,et al.  Generating a Tolerogenic Cell Therapy Knowledge Graph from Literature , 2017, Front. Immunol..

[8]  Jeffrey Heer,et al.  SpanningAspectRatioBank Easing FunctionS ArrayIn ColorIn Date Interpolator MatrixInterpola NumObjecPointI Rectang ISchedu Parallel Pause Scheduler Sequen Transition Transitioner Transiti Tween Co DelimGraphMLCon IData JSONCon DataField DataSc Dat DataSource Data DataUtil DirtySprite LineS RectSprite , 2011 .

[9]  George Hripcsak,et al.  Automated acquisition of disease drug knowledge from biomedical and clinical documents: an initial study. , 2008, Journal of the American Medical Informatics Association : JAMIA.

[10]  Halil Kilicoglu,et al.  Using semantic predications to uncover drug-drug interactions in clinical data , 2014, J. Biomed. Informatics.

[11]  Gang Pan,et al.  Semantic Health Knowledge Graph: Semantic Integration of Heterogeneous Medical Knowledge and Services , 2017, BioMed research international.

[12]  Carrie Printz.  Two‐drug combination benefits patients with chronic lymphocytic leukemia , 2020, Cancer.

[13]  Ming-Wei Chang,et al.  BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding , 2019, NAACL.

[14]  Christopher S. G. Khoo,et al.  Automatic identification of treatment relations for medical ontology learning : an exploratory study , 2004 .

[15]  Rong Xu,et al.  Large-scale extraction of accurate drug-disease treatment pairs from biomedical literature for drug repurposing , 2013, BMC Bioinformatics.

[16]  Halil Kilicoglu,et al.  Constructing a semantic predication gold standard from the biomedical literature , 2011, BMC Bioinformatics.

[17]  Wei Zheng,et al.  Drug combination therapy increases successful drug repositioning. , 2016, Drug discovery today.

[18]  Serguei V. S. Pakhomov,et al.  Evaluating active learning methods for annotating semantic predications , 2018, JAMIA open.

[19]  DAISUKE OTA,et al.  Phase I Study of Combination Therapy With Weekly Nanoparticle Albumin-bound Paclitaxel and Cyclophosphamide in Metastatic Breast Cancer Patients , 2019, AntiCancer Research.

[20]  Halil Kilicoglu,et al.  Assigning factuality values to semantic relations extracted from biomedical research literature , 2017, PloS one.

[21]  Halil Kilicoglu,et al.  Towards a characterization of apparent contradictions in the biomedical literature using context analysis , 2019, J. Biomed. Informatics.

[22]  Juan Liu,et al.  Predicting drug-disease interactions by semi-supervised graph cut algorithm and three-layer data integration , 2017, BMC Medical Genomics.

[23]  Ying Liu,et al.  Using SemRep to Label Semantic Relations Extracted from Clinical Text , 2012, AMIA.

[24]  Mengnan Zhao,et al.  Drug Repositioning to Accelerate Drug Development Using Social Media Data: Computational Study on Parkinson Disease , 2018, Journal of medical Internet research.

[25]  Aida Bchir,et al.  Extraction of drug-disease relations from MEDLINE abstracts , 2013, 2013 World Congress on Computer and Information Technology (WCCIT).

[26]  Godefridus J. Peters,et al.  Combination therapies of Artemisinin and its derivatives as a viable approach for future cancer treatment. , 2019, Current pharmaceutical design.

[27]  Shixian Ning,et al.  Knowledge-guided convolutional networks for chemical-disease relation extraction , 2019, BMC Bioinformatics.

[28]  Erik M. van Mulligen,et al.  Automated extraction of potential migraine biomarkers using a semantic graph , 2017, J. Biomed. Informatics.

[29]  G. Almouzni,et al.  Combining epigenetic drugs with other therapies for solid tumours — past lessons and future promise , 2019, Nature Reviews Clinical Oncology.

[30]  Halil Kilicoglu,et al.  SemMedDB: a PubMed-scale repository of biomedical semantic predications , 2012, Bioinform..

[31]  David Sontag,et al.  Learning a Health Knowledge Graph from Electronic Medical Records , 2017, Scientific Reports.