Dataset Information

Neuro-symbolic representation learning on biological knowledge graphs.

ABSTRACT: Biological data and knowledge bases increasingly rely on Semantic Web technologies and the use of knowledge graphs for data integration, retrieval and federated queries. In the past years, feature learning methods that are applicable to graph-structured data are becoming available, but have not yet widely been applied and evaluated on structured biological knowledge. Results: We develop a novel method for feature learning on biological knowledge graphs. Our method combines symbolic methods, in particular knowledge representation using symbolic logic and automated reasoning, with neural networks to generate embeddings of nodes that encode for related information within knowledge graphs. Through the use of symbolic logic, these embeddings contain both explicit and implicit information. We apply these embeddings to the prediction of edges in the knowledge graph representing problems of function prediction, finding candidate genes of diseases, protein-protein interactions, or drug target relations, and demonstrate performance that matches and sometimes outperforms traditional approaches based on manually crafted features. Our method can be applied to any biological knowledge graph, and will thereby open up the increasing amount of Semantic Web based knowledge bases in biology to use in machine learning and data analytics.https://github.com/bio-ontology-research-group/walking-rdf-and-owl.robert.hoehndorf@kaust.edu.sa.Supplementary data are available at Bioinformatics online.

SUBMITTER: Alshahrani M

PROVIDER: S-EPMC5860058 | biostudies-literature | 2017 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Neuro-symbolic representation learning on biological knowledge graphs.

Alshahrani Mona M Khan Mohammad Asif MA Maddouri Omar O Kinjo Akira R AR Queralt-Rosinach Núria N Hoehndorf Robert R

Bioinformatics (Oxford, England) 20170901 17

<h4>Motivation</h4>Biological data and knowledge bases increasingly rely on Semantic Web technologies and the use of knowledge graphs for data integration, retrieval and federated queries. In the past years, feature learning methods that are applicable to graph-structured data are becoming available, but have not yet widely been applied and evaluated on structured biological knowledge. Results: We develop a novel method for feature learning on biological knowledge graphs. Our method combines sym ...[more]

PMID: 28449114

Similar Datasets

Project description: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. 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. 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. 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.

Project description:BackgroundComputational analysis of complex diseases involving multiple organs requires the integration of multiple different models into a unified model. Different models are often constructed in heterogeneous formats. Thus, the integration of the models requires a standard language format that can effectively represent essential biological information. However, the previously introduced formats have limitations that prevent from adequately representing essential biological information, particularly specifications of bio-molecules and biological contexts.ResultsWe defined an XML-based markup language called context-oriented directed association markup language (CODA-ML), which better represents essential biological information. The CODA-ML has two major strengths in designating molecular specifications and biological contexts. It can cover heterogeneous entity types involved in biological events (e.g. gene/protein, compound, cellular function, disease). Molecular types of entities can have molecular specifications which include detailed information of a molecule from isoforms to modifications, enabling high-resolution representation of molecules. In addition, it can distinguish biological events that vary depending on different biological contexts such as cell types or disease conditions. Especially representation of inter-cellular events as well as intra-cellular events is available. These two major strengths can resolve contradictory associations when different models are integrated into one unified model, which improves the accuracy of the model.ConclusionsWith the CODA-ML, diverse models such as signaling pathways, metabolic pathways, and gene regulatory pathways can be represented in a unified language format. Heterogeneous entity types can be covered by the CODA-ML, thus it enables detailed description for the mechanisms of diseases or drugs from multiple perspectives (e.g., molecule, function or disease). The CODA-ML is expected to help integrate different models into one systemic model in an efficient and effective. The unified model can be used to perform computational analysis not only for cancer but also for other complex diseases involving multiple organs beyond a single cell.

Project description:Proper expression of the genes plays a vital role in the function of an organism. Recent advancements in DNA microarray technology allow for monitoring the expression level of thousands of genes. One of the important tasks in this context is to understand the underlying mechanisms of gene regulation. Recently, researchers have focused on identifying local DNA elements, or motifs to infer the relation between the expression and the nucleotide sequence of the gene. This study proposes a novel data adaptive representation approach for supervised learning to predict the response associated with the biological sequences. Biological sequences such as DNA and protein are a class of categorical sequences. In machine learning, categorical sequences are generally mapped to a lower dimensional representation for learning tasks to avoid problems with high dimensionality. The proposed method, namely SW-RF (sliding window-random forest), is a feature-based approach requiring two main steps to learn a representation for categorical sequences. In the first step, each sequence is represented by overlapping subsequences of constant length. Then a tree-based learner on this representation is trained to obtain a bag-of-words like representation which is the frequency of subsequences on the terminal nodes of the tree for each sequence. After representation learning, any classifier can be trained on the learned representation. A lasso logistic regression is trained on the learned representation to facilitate the identification of important patterns for the classification task. Our experiments show that proposed approach provides significantly better results in terms of accuracy on both synthetic data and DNA promoter sequence data. Moreover, a common problem for microarray datasets, namely missing values, is handled efficiently by the tree learners in SW-RF. Although the focus of this paper is on biological sequences, SW-RF is flexible in handling any categorical sequence data from different applications.

Dataset Information

Neuro-symbolic representation learning on biological knowledge graphs.

Publications

Neuro-symbolic representation learning on biological knowledge graphs.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets