Biological Expression Language Research Articles

A natural language processing system for the efficient updating of highly curated pathophysiology mechanism knowledge graphs

BackgroundBiomedical knowledge graphs (KG) have become crucial for describing biological findings in a structured manner. To keep up with the constantly changing flow of knowledge, their embedded information must be regularly updated with the latest findings. Natural language processing (NLP) has created new possibilities for automating this upkeep by facilitating information extraction from free text. However, due to annotated and labeled biomedical data limitations, the development of completely autonomous information extraction systems remains a substantial scientific and technological hurdle. This study aims to explore methodologies best suited to support the automatic extraction of causal relationships from biomedical literature with the aim of regular and rapid updating of disease-specific pathophysiology mechanism KGs. MethodsOur proposed approach first searches and retrieves PubMed abstracts using the desired terms and keywords. The extension corpora are then passed through the NLP pipeline for automatic information extraction. We then identify triples representing cause-and-effect relationships and encode this content using the Biological Expression Language (BEL). Finally, domain experts perform an analysis of the completeness, relevance, accuracy, and novelty of the extracted triples. ResultsIn our test scenario, which is focused on the KG regarding the phosphorylation of the Tau protein, our pipeline successfully contributed novel data, which was then subsequently used to update the KG leading to the identification of six additional upstream regulators of Tau phosphorylation. ConclusionHere, it is demonstrated that the NLP-based workflow we created is capable of rapidly updating pathophysiology mechanism graphs. As a result, production-scale, semi-automated updating of pre-existing, curated mechanism graphs is enabled.

Context mining and graph queries on giant biomedical knowledge graphs

Contextual information is widely considered for NLP and knowledge discovery in life sciences since it highly influences the exact meaning of natural language. The scientific challenge is not only to extract such context data, but also to store this data for further query and discovery approaches. Classical approaches use RDF triple stores, which have serious limitations. Here, we propose a multiple step knowledge graph approach using labeled property graphs based on polyglot persistence systems to utilize context data for context mining, graph queries, knowledge discovery and extraction. We introduce the graph-theoretic foundation for a general context concept within semantic networks and show a proof of concept based on biomedical literature and text mining. Our test system contains a knowledge graph derived from the entirety of PubMed and SCAIView data and is enriched with text mining data and domain-specific language data using Biological Expression Language. Here, context is a more general concept than annotations. This dense graph has more than 71M nodes and 850M relationships. We discuss the impact of this novel approach with 27 real-world use cases represented by graph queries. Storing and querying a giant knowledge graph as a labeled property graph is still a technological challenge. Here, we demonstrate how our data model is able to support the understanding and interpretation of biomedical data. We present several real-world use cases that utilize our massive, generated knowledge graph derived from PubMed data and enriched with additional contextual data. Finally, we show a working example in context of biologically relevant information using SCAIView.

Causal biological network models for reactive astrogliosis: a systems approach to neuroinflammation

Astrocytes play a central role in the neuroimmune response by responding to CNS pathologies with diverse molecular and morphological changes during the process of reactive astrogliosis. Here, we used a computational biological network model and mathematical algorithms that allow the interpretation of high-throughput transcriptomic datasets in the context of known biology to study reactive astrogliosis. We gathered available mechanistic information from the literature into a comprehensive causal biological network (CBN) model of astrocyte reactivity. The CBN model was built in the Biological Expression Language, which is both human-readable and computable. We characterized the CBN with a network analysis of highly connected nodes and demonstrated that the CBN captures relevant astrocyte biology. Subsequently, we used the CBN and transcriptomic data to identify key molecular pathways driving the astrocyte phenotype in four CNS pathologies: samples from mouse models of lipopolysaccharide-induced endotoxemia, Alzheimer’s disease, and amyotrophic lateral sclerosis; and samples from multiple sclerosis patients. The astrocyte CBN provides a new tool to identify causal mechanisms and quantify astrogliosis based on transcriptomic data.

Identifying neuroinflammation in omics datasets with a computational biological network model of reactive astrogliosis

Neuroimmune responses by astrocytes and microglia play a key role in the pathogenesis of neurodegenerative diseases. Reactive astrocytes have diverse responses to triggers and can release a wide variety of effector molecules in an insult-specific manner. Identifying key mechanisms of reactive astrogliosis in large datasets is key to finding new therapeutic targets, but the interpretation of such datasets is challenging. Causal biological network (CBNs) models assemble available biological knowledge in a structured computable format that facilitates mechanistic interpretation of molecular data in a well-defined biological context. Here, we present a CBN model of pan-reactive astrogliosis genes as a new tool to explore and quantify the extent of reactive astrogliosis based on transcriptomic data and identify key molecular drivers of the phenotype. Mechanistic knowledge from 133 original publications on astrogliosis were used to build this CBN model. The statements were translated into computational statements using the Biological Expression Language and assembled into a CBN. Biological entities are represented by molecular nodes connected with edges that indicate the causal relationship between these nodes. A collection of these causal relationships were assembled to form a CBN. Information about gene expression regulation by a subset of the nodes was used to build a second scorable layer to infer the activity of the CBN nodes in transcriptomic datasets. The extent of perturbation of the astrogliosis CBN in transcriptomic datasets was quantified with a network perturbation amplitude algorithm. The astrogliosis CBN model is composed of 879 nodes and 1687 edges. Highly connected nodes within the network included the S1PR family of Sphingosine 1-phosphate receptors, extracellular matrix proteins, inflammatory regulator nuclear factor NF-κB complex and the glial fibrillary acidic protein. Assessment of transcriptomic dataset GSE75246 in the context of the CBN showed increased astrogliosis in astrocytes from LPS-treated mice compared with vehicle-treated mouse astrocytes. Increased signaling of the toll-like receptor 4, interferons and interleukin-6 family cytokines and decreased acetylcholine esterase and histone deacetylase activity were inferred as key drivers of LPS-induced astrogliosis in this dataset. The astrogliosis CBN is a novel tool to quantify astrogliosis in transcriptomic datasets and identify insult-specific molecular drivers of the phenotype.

Systems biology approach highlights mechanistic differences between Crohn\u2019s disease and ulcerative colitis

The molecular mechanisms of IBD have been the subject of intensive exploration. We, therefore, assembled the available information into a suite of causal biological network models, which offer comprehensive visualization of the processes underlying IBD. Scientific text was curated by using Biological Expression Language (BEL) and compiled with OpenBEL 3.0.0. Network properties were analysed by Cytoscape. Network perturbation amplitudes were computed to score the network models with transcriptomic data from public data repositories. The IBD network model suite consists of three independent models that represent signalling pathways that contribute to IBD. In the “intestinal permeability” model, programmed cell death factors were downregulated in CD and upregulated in UC. In the “inflammation” model, PPARG, IL6, and IFN-associated pathways were prominent regulatory factors in both diseases. In the “wound healing” model, factors promoting wound healing were upregulated in CD and downregulated in UC. Scoring of publicly available transcriptomic datasets onto these network models demonstrated that the IBD models capture the perturbation in each dataset accurately. The IBD network model suite can provide better mechanistic insights of the transcriptional changes in IBD and constitutes a valuable tool in personalized medicine to further understand individual drug responses in IBD.

Extraction of causal relations based on SBEL and BERT model.

Extraction of causal relations between biomedical entities in the form of Biological Expression Language (BEL) poses a new challenge to the community of biomedical text mining due to the complexity of BEL statements. We propose a simplified form of BEL statements [Simplified Biological Expression Language (SBEL)] to facilitate BEL extraction and employ BERT (Bidirectional Encoder Representation from Transformers) to improve the performance of causal relation extraction (RE). On the one hand, BEL statement extraction is transformed into the extraction of an intermediate form—SBEL statement, which is then further decomposed into two subtasks: entity RE and entity function detection. On the other hand, we use a powerful pretrained BERT model to both extract entity relations and detect entity functions, aiming to improve the performance of two subtasks. Entity relations and functions are then combined into SBEL statements and finally merged into BEL statements. Experimental results on the BioCreative-V Track 4 corpus demonstrate that our method achieves the state-of-the-art performance in BEL statement extraction with F1 scores of 54.8% in Stage 2 evaluation and of 30.1% in Stage 1 evaluation, respectively. Database URL: https://github.com/grapeff/SBEL_datasets

Data-Driven Modeling of Knowledge Assemblies in Understanding Comorbidity Between Type 2 Diabetes Mellitus and Alzheimer’s Disease

Background:Recent studies have suggested comorbid association between Alzheimer’s disease (AD) and type 2 diabetes mellitus (T2DM) through identification of shared molecular mechanisms. However, the inference is pre-dominantly literature-based and lacks interpretation of pre-disposed genomic variants and transcriptomic measurables.Objective:In this study, we aim to identify shared genetic variants and dysregulated genes in AD and T2DM and explore their functional roles in the comorbidity between the diseases.Methods:The genetic variants for AD and T2DM were retrieved from GWAS catalog, GWAS central, dbSNP, and DisGeNet and subjected to linkage disequilibrium analysis. Next, shared variants were prioritized using RegulomeDB and Polyphen-2. Afterwards, a knowledge assembly embedding prioritized variants and their corresponding genes was created by mining relevant literature using Biological Expression Language. Finally, coherently perturbed genes from gene expression meta-analysis were mapped to the knowledge assembly to pinpoint biological entities and processes and depict a mechanistic link between AD and T2DM.Results:Our analysis identified four genes (i.e., ABCG1, COMT, MMP9, and SOD2) that could have dual roles in both AD and T2DM. Using cartoon representation, we have illustrated a set of causal events surrounding these genes which are associated to biological processes such as oxidative stress, insulin resistance, apoptosis and cognition.Conclusion:Our approach of using data as the driving force for unraveling disease etiologies eliminates literature bias and enables identification of novel entities that serve as the bridge between comorbid conditions.

Causal inferencing of -omics data from The Cancer Genome Atlas: Lung adenocarcinoma tumors for mechanistic disease characterization and feature engineering.

e21016 Background: Advances in high throughput measurement technologies (-omics data) have made it possible to generate high complexity, high volume data for oncology research. Researchers are often confronted many more measurements than samples (p > > > n), which poses challenges for both modeling the complexity of disease at the molecular mechanism level, and overfitting when generating predictive models with complex data. Here, we applied a prior knowledge-driven approach to characterize and classify heavy versus light smokers with lung cancer from The Cancer Genome Atlas, an open source repository that catalogs, harmonizes and hosts -omics data collected from samples generously donated from cancer patients. Methods: We applied a reverse inferencing approach to systematically interrogate RNAseq measurements from tumor and control biopsies against a knowledgebase of directed gene networks curated from published experiments. If patterns observed in the data are significantly similar to those in a network, an inference about the directional activity of that network can be made; e.g., the increased transcriptional activity of NFKB. Our library was nucleated through an open sourced knowledge graph and enhanced with updated and relevant knowledge using the open sourced Biological Expression Language framework. Directed networks were either qualitatively scored and used to build disease models, or semi-quantitatively scored and used as classification features. Results: In LUAD tumors, we detected a pattern of gene signatures which indicated a tumor stem cell-like phenotype characterized by predicted decreases in the activity of pro-differentiation factors and an increased response to hypoxia. Analysis of patients with heavy ( > 40) versus light ( < 10) pack-year burden suggested an augmented dedifferentiation profile in heavy smokers. In this example, improved classification was observed through features compression through directed network scoring compared to using individual RNA measurements selected by filtration methods. Conclusions: In-silico analysis of lung cancer patient biopsies generated hypotheses implicating stem cell signaling in tumors, and a further stratification of this signal based on patient pack year burden. Mechanistic modeling may be a useful application to the overfitting problem often encountered with -omics data in translational studies. Data from other TCGA indications can be used to evaluate the consistency of this type of approach

NPA: an R package for computing network perturbation amplitudes using gene expression data and two-layer networks

BackgroundHigh-throughput gene expression technologies provide complex datasets reflecting mechanisms perturbed in an experiment, typically in a treatment versus control design. Analysis of these information-rich data can be guided based on a priori knowledge, such as networks of related proteins or genes. Assessing the response of a specific mechanism and investigating its biological basis is extremely important in systems toxicology; as compounds or treatment need to be assessed with respect to a predefined set of key mechanisms that could lead to toxicity. Two-layer networks are suitable for this task, and a robust computational methodology specifically addressing those needs was previously published.The NPA package (https://github.com/philipmorrisintl/NPA) implements the algorithm, and a data package of eight two-layer networks representing key mechanisms, such as xenobiotic metabolism, apoptosis, or epithelial immune innate activation, is provided.ResultsGene expression data from an animal study are analyzed using the package and its network models. The functionalities are implemented using R6 classes, making the use of the package seamless and intuitive. The various network responses are analyzed using the leading node analysis, and an overall perturbation, called the Biological Impact Factor, is computed.ConclusionsThe NPA package implements the published network perturbation amplitude methodology and provides a set of two-layer networks encoded in the Biological Expression Language.

PathMe: merging and exploring mechanistic pathway knowledge

BackgroundThe complexity of representing biological systems is compounded by an ever-expanding body of knowledge emerging from multi-omics experiments. A number of pathway databases have facilitated pathway-centric approaches that assist in the interpretation of molecular signatures yielded by these experiments. However, the lack of interoperability between pathway databases has hindered the ability to harmonize these resources and to exploit their consolidated knowledge. Such a unification of pathway knowledge is imperative in enhancing the comprehension and modeling of biological abstractions.ResultsHere, we present PathMe, a Python package that transforms pathway knowledge from three major pathway databases into a unified abstraction using Biological Expression Language as the pivotal, integrative schema. PathMe is complemented by a novel web application (freely available at https://pathme.scai.fraunhofer.de/) which allows users to comprehensively explore pathway crosstalk and compare areas of consensus and discrepancies.ConclusionsThis work has harmonized three major pathway databases and transformed them into a unified schema in order to gain a holistic picture of pathway knowledge. We demonstrate the utility of the PathMe framework in: i) integrating pathway landscapes at the database level, ii) comparing the degree of consensus at the pathway level, and iii) exploring pathway crosstalk and investigating consensus at the molecular level.

Hierarchical sequence labeling for extracting BEL statements from biomedical literature

BackgroundExtracting relations between bio-entities from biomedical literature is often a challenging task and also an essential step towards biomedical knowledge expansion. The BioCreative community has organized a shared task to evaluate the robustness of the causal relationship extraction algorithms in Biological Expression Language (BEL) from biomedical literature.MethodWe first map the sentence-level BEL statements in the BC-V training corpus to the corresponding text segments, thus generating hierarchically tagged training instances. A hierarchical sequence labeling model was afterwards induced from these training instances and applied to the test sentences in order to construct the BEL statements.ResultsThe experimental results on extracting BEL statements from BioCreative V Track 4 test corpus show that our method achieves promising performance with an overall F-measure of 31.6%. Furthermore, it has the potential to be enhanced by adopting more advanced machine learning approaches.ConclusionWe propose a framework for hierarchical relation extraction using hierarchical sequence labeling on the instance-level training corpus derived from the original sentence-level corpus via word alignment. Its main advantage is that we can make full use of the original training corpus to induce the sequence labelers and then apply them to the test corpus.

Construction of a Suite of Computable Biological Network Models Focused on Mucociliary Clearance in the Respiratory Tract.

Mucociliary clearance (MCC), considered as a collaboration of mucus secreted from goblet cells, the airway surface liquid layer, and the beating of cilia of ciliated cells, is the airways’ defense system against airborne contaminants. Because the process is well described at the molecular level, we gathered the available information into a suite of comprehensive causal biological network (CBN) models. The suite consists of three independent models that represent (1) cilium assembly, (2) ciliary beating, and (3) goblet cell hyperplasia/metaplasia and that were built in the Biological Expression Language, which is both human-readable and computable. The network analysis of highly connected nodes and pathways demonstrated that the relevant biology was captured in the MCC models. We also show the scoring of transcriptomic data onto these network models and demonstrate that the models capture the perturbation in each dataset accurately. This work is a continuation of our approach to use computational biological network models and mathematical algorithms that allow for the interpretation of high-throughput molecular datasets in the context of known biology. The MCC network model suite can be a valuable tool in personalized medicine to further understand heterogeneity and individual drug responses in complex respiratory diseases.

The extraction of complex relationships and their conversion to biological expression language (BEL) overview of the BioCreative VI (2017) BEL track

Knowledge of the molecular interactions of biological and chemical entities and their involvement in biological processes or clinical phenotypes is important for data interpretation. Unfortunately, this knowledge is mostly embedded in the literature in such a way that it is unavailable for automated data analysis procedures. Biological expression language (BEL) is a syntax representation allowing for the structured representation of a broad range of biological relationships. It is used in various situations to extract such knowledge and transform it into BEL networks. To support the tedious and time-intensive extraction work of curators with automated methods, we developed the BEL track within the framework of BioCreative Challenges. Within the BEL track, we provide training data and an evaluation environment to encourage the text mining community to tackle the automatic extraction of complex BEL relationships. In 2017 BioCreative VI, the 2015 BEL track was repeated with new test data. Although only minor improvements in text snippet retrieval for given statements were achieved during this second BEL task iteration, a significant increase of BEL statement extraction performance from provided sentences could be seen. The best performing system reached a 32% F-score for the extraction of complete BEL statements and with the given named entities this increased to 49%. This time, besides rule-based systems, new methods involving hierarchical sequence labeling and neural networks were applied for BEL statement extraction.

Re-curation and rational enrichment of knowledge graphs in Biological Expression Language.

The rapid accumulation of new biomedical literature not only causes curated knowledge graphs (KGs) to become outdated and incomplete, but also makes manual curation an impractical and unsustainable solution. Automated or semi-automated workflows are necessary to assist in prioritizing and curating the literature to update and enrich KGs. We have developed two workflows: one for re-curating a given KG to assure its syntactic and semantic quality and another for rationally enriching it by manually revising automatically extracted relations for nodes with low information density. We applied these workflows to the KGs encoded in Biological Expression Language from the NeuroMMSig database using content that was pre-extracted from MEDLINE abstracts and PubMed Central full-text articles using text mining output integrated by INDRA. We have made this workflow freely available at https://github.com/bel-enrichment/bel-enrichment.

Combining relation extraction with function detection for BEL statement extraction.

The BioCreative-V community proposed a challenging task of automatic extraction of causal relation network in Biological Expression Language (BEL) from the biomedical literature. Previous studies on this task largely used models induced from other related tasks and then transformed intermediate structures to BEL statements, which left the given training corpus unexplored. To make full use of the BEL training corpus, in this work, we propose a deep learning-based approach to extract BEL statements. Specifically, we decompose the problem into two subtasks: entity relation extraction and entity function detection. First, two attention-based bidirectional long short-term memory networks models are used to extract entity relation and entity function, respectively. Then entity relation and their functions are combined into a BEL statement. In order to boost the overall performance, a strategy of threshold filtering is applied to improve the precision of identified entity functions. We evaluate our approach on the BioCreative-V Track 4 corpus with or without gold entities. The experimental results show that our method achieves the state-of-the-art performance with an overall F1-measure of 46.9% in stage 2 and 21.3% in stage 1, respectively.

BEL2ABM: agent-based simulation of static models in Biological Expression Language.

SummaryWhile cause-and-effect knowledge assembly models encoded in Biological Expression Language are able to support generation of mechanistic hypotheses, they are static and limited in their ability to encode temporality. Here, we present BEL2ABM, a software for producing continuous, dynamic, executable agent-based models from BEL templates.Availability and implementationThe tool has been developed in Java and NetLogo. Code, data and documentation are available under the Apache 2.0 License at https://github.com/pybel/bel2abm.Supplementary information Supplementary data are available at Bioinformatics online.

BEL Commons: an environment for exploration and analysis of networks encoded in Biological Expression Language.

The rapid accumulation of knowledge in the field of systems and networks biology during recent years requires complex, but user-friendly and accessible web applications that allow from visualization to complex algorithmic analysis. While several web applications exist with various focuses on creation, revision, curation, storage, integration, collaboration, exploration, visualization and analysis, many of these services remain disjoint and have yet to be packaged into a cohesive environment.Here, we present BEL Commons: an integrative knowledge discovery environment for networks encoded in the Biological Expression Language (BEL). Users can upload files in BEL to be parsed, validated, compiled and stored with fine granular permissions. After, users can summarize, explore and optionally shared their networks with the scientific community. We have implemented a query builder wizard to help users find the relevant portions of increasingly large and complex networks and a visualization interface that allows them to explore their resulting networks. Finally, we have included a dedicated analytical service for performing data-driven analysis of knowledge networks to support hypothesis generation.

A systematic approach for identifying shared mechanisms in epilepsy and its comorbidities.

Cross-sectional epidemiological studies have shown that the incidence of several nervous system diseases is more frequent in epilepsy patients than in the general population. Some comorbidities [e.g. Alzheimer’s disease (AD) and Parkinson’s disease] are also risk factors for the development of seizures; suggesting they may share pathophysiological mechanisms with epilepsy. A literature-based approach was used to identify gene overlap between epilepsy and its comorbidities as a proxy for a shared genetic basis for disease, or genetic pleiotropy, as a first effort to identify shared mechanisms. While the results identified neurological disorders as the group of diseases with the highest gene overlap, this analysis was insufficient for identifying putative common mechanisms shared across epilepsy and its comorbidities. This motivated the use of a dedicated literature mining and knowledge assembly approach in which a cause-and-effect model of epilepsy was captured with Biological Expression Language. After enriching the knowledge assembly with information surrounding epilepsy, its risk factors, its comorbidities, and anti-epileptic drugs, a novel comparative mechanism enrichment approach was used to propose several downstream effectors (including the GABA receptor, GABAergic pathways, etc.) that could explain the therapeutic effects carbamazepine in both the contexts of epilepsy and AD. We have made the Epilepsy Knowledge Assembly available at https://www.scai.fraunhofer.de/content/dam/scai/de/downloads/bioinformatik/epilepsy.bel and queryable through NeuroMMSig at http://neurommsig.scai.fraunhofer.de. The source code used for analysis and tutorials for reproduction are available on GitHub at https://github.com/cthoyt/epicom.

Integrated approaches in diagnostics and therapy of allergic diseases

Allergic diseases are a serious problem in both developed and developing countries. Based on the World Health Organization data over 30-40% of the population have one or more allergic diseases. According to forecasts, by 2050, up to 4 billion people in the world will suffer from asthma, allergic rhinitis or atopic dermatitis. Solving problems related to the complexity of differential diagnosis, false positive and false negative results of clinical and laboratory studies, genetic characteristics of patients and many others, can be realized by integrating approaches of bioinformatics and systems biomedicine based on massive databases of experimental studies on one side and, on the other - on advanced technologies of genotyping and detection of biomarkers. The review analyzes the main resources of international databases on allergens, which help to determine the main characteristics of allergens: molecular weight, epitopes, cross reactivity, geographical prevalence, availability allergens in food. Different approaches are considered in the systematization of data obtained in the study of the genome, transcriptome, microbiome, comparison of data obtained from healthy donors and patients with allergic diseases, genetic mutations, transcriptome and microbiome profiles that cause severe course of allergic diseases. Several ways of depicting relationships in the construction of signaling networks (KEGG, sbvIMPROVER, Cyto scape) are shown, both on the basis of direct influence (KEGG, Cytoscape) and on the basis of OpenBEL - the open-access biological expression language - sbvIMPROVER, capable of displaying complex semantic links between components of the system under consideration.

Comorbidity Analysis between Alzheimer's Disease and Type 2 Diabetes Mellitus (T2DM) Based on Shared Pathways and the Role of T2DM Drugs.

Background:Various studies suggest a comorbid association between Alzheimer’s disease (AD) and type 2 diabetes mellitus (T2DM) indicating that there could be shared underlying pathophysiological mechanisms.Objective:This study aims to systematically model relevant knowledge at the molecular level to find a mechanistic rationale explaining the existing comorbid association between AD and T2DM.Method:We have used a knowledge-based modeling approach to build two network models for AD and T2DM using Biological Expression Language (BEL), which is capable of capturing and representing causal and correlative relationships at both molecular and clinical levels from various knowledge resources.Results:Using comparative analysis, we have identified several putative “shared pathways”. We demonstrate, at a mechanistic level, how the insulin signaling pathway is related to other significant AD pathways such as the neurotrophin signaling pathway, PI3K/AKT signaling, MTOR signaling, and MAPK signaling and how these pathways do cross-talk with each other both in AD and T2DM. In addition, we present a mechanistic hypothesis that explains both favorable and adverse effects of the anti-diabetic drug metformin in AD.Conclusion:The two computable models introduced here provide a powerful framework to identify plausible mechanistic links shared between AD and T2DM and thereby identify targeted pathways for new therapeutics. Our approach can also be used to provide mechanistic answers to the question of why some T2DM treatments seem to increase the risk of AD.