Identifying disease-gene associations using a convolutional neural network-based model by embedding a biological knowledge graph with entity descriptions

Petagraph: A large-scale unifying knowledge graph framework for integrating biomolecular and biomedical data

Over the past decade, there has been substantial growth in both the quantity and complexity of available biomedical data. In order to more efficiently harness this extensive data and alleviate challenges associated with integration of multi-omics data, we developed Petagraph, a biomedical knowledge graph that encompasses over 32 million nodes and 118 million relationships. Petagraph leverages more than 180 ontologies and standards in the Unified Biomedical Knowledge Graph (UBKG) to embed millions of quantitative genomics data points. Petagraph provides a cohesive data environment that enables users to efficiently analyze, annotate, and discern relationships within and across complex multi-omics datasets supported by UBKG's annotation scaffold. We demonstrate how queries on Petagraph can generate meaningful results across various research contexts and use cases.

Predicting gene disease associations with knowledge graph embeddings for diseases with curtailed information

Knowledge graph embeddings (KGE) are a powerful technique used in the biomedical domain to represent biological knowledge in a low dimensional space. However, a deep understanding of these methods is still missing, and, in particular, regarding their applications to prioritize genes associated with complex diseases with reduced genetic information. In this contribution, we built a knowledge graph (KG) by integrating heterogeneous biomedical data and generated KGE by implementing state-of-the-art methods, and two novel algorithms: Dlemb and BioKG2vec. Extensive testing of the embeddings with unsupervised clustering and supervised methods showed that KGE can be successfully implemented to predict genes associated with diseases and that our novel approaches outperform most existing algorithms in both scenarios. Our findings underscore the significance of data quality, preprocessing, and integration in achieving accurate predictions. Additionally, we applied KGE to predict genes linked to Intervertebral Disc Degeneration (IDD) and illustrated that functions pertinent to the disease are enriched within the prioritized gene set.

Elucidating the semantics-topology trade-off for knowledge inference-based pharmacological discovery

Leveraging AI for synthesizing the deluge of biomedical knowledge has great potential for pharmacological discovery with applications including developing new therapeutics for untreated diseases and repurposing drugs as emergent pandemic treatments. Creating knowledge graph representations of interacting drugs, diseases, genes, and proteins enables discovery via embedding-based ML approaches and link prediction. Previously, it has been shown that these predictive methods are susceptible to biases from network structure, namely that they are driven not by discovering nuanced biological understanding of mechanisms, but based on high-degree hub nodes. In this work, we study the confounding effect of network topology on biological relation semantics by creating an experimental pipeline of knowledge graph semantic and topological perturbations. We show that the drop in drug repurposing performance from ablating meaningful semantics increases by 21% and 38% when mitigating topological bias in two networks. We demonstrate that new methods for representing knowledge and inferring new knowledge must be developed for making use of biomedical semantics for pharmacological innovation, and we suggest fruitful avenues for their development.

KDGene: knowledge graph completion for disease gene prediction using interactional tensor decomposition

The accurate identification of disease-associated genes is crucial for understanding the molecular mechanisms underlying various diseases. Most current methods focus on constructing biological networks and utilizing machine learning, particularly deep learning, to identify disease genes. However, these methods overlook complex relations among entities in biological knowledge graphs. Such information has been successfully applied in other areas of life science research, demonstrating their effectiveness. Knowledge graph embedding methods can learn the semantic information of different relations within the knowledge graphs. Nonetheless, the performance of existing representation learning techniques, when applied to domain-specific biological data, remains suboptimal. To solve these problems, we construct a biological knowledge graph centered on diseases and genes, and develop an end-to-end knowledge graph completion framework for disease gene prediction using interactional tensor decomposition named KDGene. KDGene incorporates an interaction module that bridges entity and relation embeddings within tensor decomposition, aiming to improve the representation of semantically similar concepts in specific domains and enhance the ability to accurately predict disease genes. Experimental results show that KDGene significantly outperforms state-of-the-art algorithms, whether existing disease gene prediction methods or knowledge graph embedding methods for general domains. Moreover, the comprehensive biological analysis of the predicted results further validates KDGene's capability to accurately identify new candidate genes. This work proposes a scalable knowledge graph completion framework to identify disease candidate genes, from which the results are promising to provide valuable references for further wet experiments. Data and source codes are available at https://github.com/2020MEAI/KDGene.

A new model construction based on the knowledge graph for mining elite polyphenotype genes in crops

Identifying polyphenotype genes that simultaneously regulate important agronomic traits (e.g., plant height, yield, and disease resistance) is critical for developing novel high-quality crop varieties. Predicting the associations between genes and traits requires the organization and analysis of multi-dimensional scientific data. The existing methods for establishing the relationships between genomic data and phenotypic data can only elucidate the associations between genes and individual traits. However, there are relatively few methods for detecting elite polyphenotype genes. In this study, a knowledge graph for traits regulating-genes was constructed by collecting data from the PubMed database and eight other databases related to the staple food crops rice, maize, and wheat as well as the model plant Arabidopsis thaliana. On the basis of the knowledge graph, a model for predicting traits regulating-genes was constructed by combining the data attributes of the gene nodes and the topological relationship attributes of the gene nodes. Additionally, a scoring method for predicting the genes regulating specific traits was developed to screen for elite polyphenotype genes. A total of 125,591 nodes and 547,224 semantic relationships were included in the knowledge graph. The accuracy of the knowledge graph-based model for predicting traits regulating-genes was 0.89, the precision rate was 0.91, the recall rate was 0.96, and the F1 value was 0.94. Moreover, 4,447 polyphenotype genes for 31 trait combinations were identified, among which the rice polyphenotype gene IPA1 and the A. thaliana polyphenotype gene CUC2 were verified via a literature search. Furthermore, the wheat gene TraesCS5A02G275900 was revealed as a potential polyphenotype gene that will need to be further characterized. Meanwhile, the result of venn diagram analysis between the polyphenotype gene datasets (consists of genes that are predicted by our model) and the transcriptome gene datasets (consists of genes that were differential expression in response to disease, drought or salt) showed approximately 70% and 54% polyphenotype genes were identified in the transcriptome datasets of Arabidopsis and rice, respectively. The application of the model driven by knowledge graph for predicting traits regulating-genes represents a novel method for detecting elite polyphenotype genes.

APDDD: Animal parasitic diseases and drugs database

Animal parasitic diseases not only have an economic impact, but also have serious social and public health impacts. Although antiparasitic drugs can treat these diseases, it seems difficult for users to comprehensively utilize the information, due to incomplete and difficult data collection. Thus, there is an urgent need to establish a comprehensive database, that includes parasitic diseases and related drugs. In this paper, we develop a knowledge database dedicated to collecting and analyzing animal parasitic diseases and related drugs, named Animal Parasitic Diseases and Drugs Database (APDDD). The current version of APDDD includes animal parasitic disease data of 8 major parasite classifications that cause common parasitic diseases and 96 subclass samples mined from many literature and authoritative books, as well as 182 antiparasitic drugs. Furthermore, we utilized APDDD data to add a knowledge graph representing the relationships between parasitic diseases, drugs, and the targeted gene of drugs acting on parasites. We hope that APDDD will become a good database for animal parasitic diseases and antiparasitic drugs research and that users can gain a more intuitive understanding of the relationships between parasitic diseases, drugs, and targeted genes through the knowledge graph.

BioBLP: a modular framework for learning on multimodal biomedical knowledge graphs

BACKGROUND

Knowledge graphs (KGs) are an important tool for representing complex relationships between entities in the biomedical domain. Several methods have been proposed for learning embeddings that can be used to predict new links in such graphs. Some methods ignore valuable attribute data associated with entities in biomedical KGs, such as protein sequences, or molecular graphs. Other works incorporate such data, but assume that entities can be represented with the same data modality. This is not always the case for biomedical KGs, where entities exhibit heterogeneous modalities that are central to their representation in the subject domain.

OBJECTIVE

We aim to understand how to incorporate multimodal data into biomedical KG embeddings, and analyze the resulting performance in comparison with traditional methods. We propose a modular framework for learning embeddings in KGs with entity attributes, that allows encoding attribute data of different modalities while also supporting entities with missing attributes. We additionally propose an efficient pretraining strategy for reducing the required training runtime. We train models using a biomedical KG containing approximately 2 million triples, and evaluate the performance of the resulting entity embeddings on the tasks of link prediction, and drug-protein interaction prediction, comparing against methods that do not take attribute data into account.

RESULTS

In the standard link prediction evaluation, the proposed method results in competitive, yet lower performance than baselines that do not use attribute data. When evaluated in the task of drug-protein interaction prediction, the method compares favorably with the baselines. Further analyses show that incorporating attribute data does outperform baselines over entities below a certain node degree, comprising approximately 75% of the diseases in the graph. We also observe that optimizing attribute encoders is a challenging task that increases optimization costs. Our proposed pretraining strategy yields significantly higher performance while reducing the required training runtime.

CONCLUSION

BioBLP allows to investigate different ways of incorporating multimodal biomedical data for learning representations in KGs. With a particular implementation, we find that incorporating attribute data does not consistently outperform baselines, but improvements are obtained on a comparatively large subset of entities below a specific node-degree. Our results indicate a potential for improved performance in scientific discovery tasks where understudied areas of the KG would benefit from link prediction methods.

Speos: an ensemble graph representation learning framework to predict core gene candidates for complex diseases

Understanding phenotype-to-genotype relationships is a grand challenge of 21st century biology with translational implications. The recently proposed "omnigenic" model postulates that effects of genetic variation on traits are mediated by core-genes and -proteins whose activities mechanistically influence the phenotype, whereas peripheral genes encode a regulatory network that indirectly affects phenotypes via core gene products. Here, we develop a positive-unlabeled graph representation-learning ensemble-approach based on a nested cross-validation to predict core-like genes for diverse diseases using Mendelian disorder genes for training. Employing mouse knockout phenotypes for external validations, we demonstrate that core-like genes display several key properties of core genes: Mouse knockouts of genes corresponding to our most confident predictions give rise to relevant mouse phenotypes at rates on par with the Mendelian disorder genes, and all candidates exhibit core gene properties like transcriptional deregulation in disease and loss-of-function intolerance. Moreover, as predicted for core genes, our candidates are enriched for drug targets and druggable proteins. In contrast to Mendelian disorder genes the new core-like genes are enriched for druggable yet untargeted gene products, which are therefore attractive targets for drug development. Interpretation of the underlying deep learning model suggests plausible explanations for our core gene predictions in form of molecular mechanisms and physical interactions. Our results demonstrate the potential of graph representation learning for the interpretation of biological complexity and pave the way for studying core gene properties and future drug development.

RelCurator: a text mining-based curation system for extracting gene-phenotype relationships specific to neurodegenerative disorders

BACKGROUND

The identification of gene-phenotype relationships is important in medical genetics as it serves as a basis for precision medicine. However, most of the gene-phenotype relationship data are buried in the biomedical literature in textual form.

OBJECTIVE

We propose RelCurator, a curation system that extracts sentences including both gene and phenotype entities related to specific disease categories from PubMed articles, provides rich additional information such as entity taggings, and predictions of gene-phenotype relationships.

METHODS

We targeted neurodegenerative disorders and developed a deep learning model using Bidirectional Gated Recurrent Unit (BiGRU) networks and BioWordVec word embeddings for predicting gene-phenotype relationships from biomedical texts. The prediction model is trained with more than 130,000 labeled PubMed sentences including gene and phenotype entities, which are related to or unrelated to neurodegenerative disorders.

RESULTS

We compared the performance of our deep learning model with those of Bidirectional Encoder Representations from Transformers (BERT), Support Vector Machine (SVM), and simple Recurrent Neural Network (simple RNN) models. Our model performed better with an F1-score of 0.96. Furthermore, the evaluation done using a few curation cases in the real scenario showed the effectiveness of our work. Therefore, we conclude that RelCurator can identify not only new causative genes, but also new genes associated with neurodegenerative disorders' phenotype.

CONCLUSION

RelCurator is a user-friendly method for accessing deep learning-based supporting information and a concise web interface to assist curators while browsing the PubMed articles. Our curation process represents an important and broadly applicable improvement to the state of the art for the curation of gene-phenotype relationships.

Healthcare knowledge graph construction: A systematic review of the state-of-the-art, open issues, and opportunities

The incorporation of data analytics in the healthcare industry has made significant progress, driven by the demand for efficient and effective big data analytics solutions. Knowledge graphs (KGs) have proven utility in this arena and are rooted in a number of healthcare applications to furnish better data representation and knowledge inference. However, in conjunction with a lack of a representative KG construction taxonomy, several existing approaches in this designated domain are inadequate and inferior. This paper is the first to provide a comprehensive taxonomy and a bird's eye view of healthcare KG construction. Additionally, a thorough examination of the current state-of-the-art techniques drawn from academic works relevant to various healthcare contexts is carried out. These techniques are critically evaluated in terms of methods used for knowledge extraction, types of the knowledge base and sources, and the incorporated evaluation protocols. Finally, several research findings and existing issues in the literature are reported and discussed, opening horizons for future research in this vibrant area.

Effective holistic characterization of small molecule effects using heterogeneous biological networks

The two most common reasons for attrition in therapeutic clinical trials are efficacy and safety. We integrated heterogeneous data to create a human interactome network to comprehensively describe drug behavior in biological systems, with the goal of accurate therapeutic candidate generation. The Computational Analysis of Novel Drug Opportunities (CANDO) platform for shotgun multiscale therapeutic discovery, repurposing, and design was enhanced by integrating drug side effects, protein pathways, protein-protein interactions, protein-disease associations, and the Gene Ontology, and complemented with its existing drug/compound, protein, and indication libraries. These integrated networks were reduced to a "multiscale interactomic signature" for each compound that describe its functional behavior as vectors of real values. These signatures are then used for relating compounds to each other with the hypothesis that similar signatures yield similar behavior. Our results indicated that there is significant biological information captured within our networks (particularly via side effects) which enhance the performance of our platform, as evaluated by performing all-against-all leave-one-out drug-indication association benchmarking as well as generating novel drug candidates for colon cancer and migraine disorders corroborated via literature search. Further, drug impacts on pathways derived from computed compound-protein interaction scores served as the features for a random forest machine learning model trained to predict drug-indication associations, with applications to mental disorders and cancer metastasis highlighted. This interactomic pipeline highlights the ability of Computational Analysis of Novel Drug Opportunities to accurately relate drugs in a multitarget and multiscale context, particularly for generating putative drug candidates using the information gleaned from indirect data such as side effect profiles and protein pathway information.