Knowledge graphs in psychiatric research: Potential applications and future perspectives

BACKGROUND

Knowledge graphs (KGs) remain an underutilized tool in the field of psychiatric research. In the broader biomedical field KGs are already a significant tool mainly used as knowledge database or for novel relation detection between biomedical entities. This review aims to outline how KGs would further research in the field of psychiatry in the age of Artificial Intelligence (AI) and Large Language Models (LLMs).

METHODS

We conducted a thorough literature review across a spectrum of scientific fields ranging from computer science and knowledge engineering to bioinformatics. The literature reviewed was taken from PubMed, Semantic Scholar and Google Scholar searches including terms such as "Psychiatric Knowledge Graphs", "Biomedical Knowledge Graphs", "Knowledge Graph Machine Learning Applications", "Knowledge Graph Applications for Biomedical Sciences". The resulting publications were then assessed and accumulated in this review regarding their possible relevance to future psychiatric applications.

RESULTS

A multitude of papers and applications of KGs in associated research fields that are yet to be utilized in psychiatric research was found and outlined in this review. We create a thorough recommendation for other computational researchers regarding use-cases of these KG applications in psychiatry.

CONCLUSION

This review illustrates use-cases of KG-based research applications in biomedicine and beyond that may aid in elucidating the complex biology of psychiatric illness and open new routes for developing innovative interventions. We conclude that there is a wealth of opportunities for KG utilization in psychiatric research across a variety of application areas including biomarker discovery, patient stratification and personalized medicine approaches.

Prioritizing genomic variants through neuro-symbolic, knowledge-enhanced learning

MOTIVATION

Whole-exome and genome sequencing have become common tools in diagnosing patients with rare diseases. Despite their success, this approach leaves many patients undiagnosed. A common argument is that more disease variants still await discovery, or the novelty of disease phenotypes results from a combination of variants in multiple disease-related genes. Interpreting the phenotypic consequences of genomic variants relies on information about gene functions, gene expression, physiology, and other genomic features. Phenotype-based methods to identify variants involved in genetic diseases combine molecular features with prior knowledge about the phenotypic consequences of altering gene functions. While phenotype-based methods have been successfully applied to prioritizing variants, such methods are based on known gene-disease or gene-phenotype associations as training data and are applicable to genes that have phenotypes associated, thereby limiting their scope. In addition, phenotypes are not assigned uniformly by different clinicians, and phenotype-based methods need to account for this variability.

RESULTS

We developed an Embedding-based Phenotype Variant Predictor (EmbedPVP), a computational method to prioritize variants involved in genetic diseases by combining genomic information and clinical phenotypes. EmbedPVP leverages a large amount of background knowledge from human and model organisms about molecular mechanisms through which abnormal phenotypes may arise. Specifically, EmbedPVP incorporates phenotypes linked to genes, functions of gene products, and the anatomical site of gene expression, and systematically relates them to their phenotypic effects through neuro-symbolic, knowledge-enhanced machine learning. We demonstrate EmbedPVP's efficacy on a large set of synthetic genomes and genomes matched with clinical information.

AVAILABILITY AND IMPLEMENTATION

EmbedPVP and all evaluation experiments are freely available at https://github.com/bio-ontology-research-group/EmbedPVP.

Computational Approaches to Predict Protein-Protein Interactions in Crowded Cellular Environments

Investigating protein-protein interactions is crucial for understanding cellular biological processes because proteins often function within molecular complexes rather than in isolation. While experimental and computational methods have provided valuable insights into these interactions, they often overlook a critical factor: the crowded cellular environment. This environment significantly impacts protein behavior, including structural stability, diffusion, and ultimately the nature of binding. In this review, we discuss theoretical and computational approaches that allow the modeling of biological systems to guide and complement experiments and can thus significantly advance the investigation, and possibly the predictions, of protein-protein interactions in the crowded environment of cell cytoplasm. We explore topics such as statistical mechanics for lattice simulations, hydrodynamic interactions, diffusion processes in high-viscosity environments, and several methods based on molecular dynamics simulations. By synergistically leveraging methods from biophysics and computational biology, we review the state of the art of computational methods to study the impact of molecular crowding on protein-protein interactions and discuss its potential revolutionizing effects on the characterization of the human interactome.

Explaining protein-protein interactions with knowledge graph-based semantic similarity

The application of artificial intelligence and machine learning methods for several biomedical applications, such as protein-protein interaction prediction, has gained significant traction in recent decades. However, explainability is a key aspect of using machine learning as a tool for scientific discovery. Explainable artificial intelligence approaches help clarify algorithmic mechanisms and identify potential bias in the data. Given the complexity of the biomedical domain, explanations should be grounded in domain knowledge which can be achieved by using ontologies and knowledge graphs. These knowledge graphs express knowledge about a domain by capturing different perspectives of the representation of real-world entities. However, the most popular way to explore knowledge graphs with machine learning is through using embeddings, which are not explainable. As an alternative, knowledge graph-based semantic similarity offers the advantage of being explainable. Additionally, similarity can be computed to capture different semantic aspects within the knowledge graph and increasing the explainability of predictive approaches. We propose a novel method to generate explainable vector representations, KGsim2vec, that uses aspect-oriented semantic similarity features to represent pairs of entities in a knowledge graph. Our approach employs a set of machine learning models, including decision trees, genetic programming, random forest and eXtreme gradient boosting, to predict relations between entities. The experiments reveal that considering multiple semantic aspects when representing the similarity between two entities improves explainability and predictive performance. KGsim2vec performs better than black-box methods based on knowledge graph embeddings or graph neural networks. Moreover, KGsim2vec produces global models that can capture biological phenomena and elucidate data biases.

Drug target prediction through deep learning functional representation of gene signatures

Many machine learning applications in bioinformatics currently rely on matching gene identities when analyzing input gene signatures and fail to take advantage of preexisting knowledge about gene functions. To further enable comparative analysis of OMICS datasets, including target deconvolution and mechanism of action studies, we develop an approach that represents gene signatures projected onto their biological functions, instead of their identities, similar to how the word2vec technique works in natural language processing. We develop the Functional Representation of Gene Signatures (FRoGS) approach by training a deep learning model and demonstrate that its application to the Broad Institute's L1000 datasets results in more effective compound-target predictions than models based on gene identities alone. By integrating additional pharmacological activity data sources, FRoGS significantly increases the number of high-quality compound-target predictions relative to existing approaches, many of which are supported by in silico and/or experimental evidence. These results underscore the general utility of FRoGS in machine learning-based bioinformatics applications. Prediction networks pre-equipped with the knowledge of gene functions may help uncover new relationships among gene signatures acquired by large-scale OMICs studies on compounds, cell types, disease models, and patient cohorts.

Partial order relation-based gene ontology embedding improves protein function prediction

Protein annotation has long been a challenging task in computational biology. Gene Ontology (GO) has become one of the most popular frameworks to describe protein functions and their relationships. Prediction of a protein annotation with proper GO terms demands high-quality GO term representation learning, which aims to learn a low-dimensional dense vector representation with accompanying semantic meaning for each functional label, also known as embedding. However, existing GO term embedding methods, which mainly take into account ancestral co-occurrence information, have yet to capture the full topological information in the GO-directed acyclic graph (DAG). In this study, we propose a novel GO term representation learning method, PO2Vec, to utilize the partial order relationships to improve the GO term representations. Extensive evaluations show that PO2Vec achieves better outcomes than existing embedding methods in a variety of downstream biological tasks. Based on PO2Vec, we further developed a new protein function prediction method PO2GO, which demonstrates superior performance measured in multiple metrics and annotation specificity as well as few-shot prediction capability in the benchmarks. These results suggest that the high-quality representation of GO structure is critical for diverse biological tasks including computational protein annotation.

Clustering rare diseases within an ontology-enriched knowledge graph

OBJECTIVE

Identifying sets of rare diseases with shared aspects of etiology and pathophysiology may enable drug repurposing. Toward that aim, we utilized an integrative knowledge graph to construct clusters of rare diseases.

MATERIALS AND METHODS

Data on 3242 rare diseases were extracted from the National Center for Advancing Translational Science Genetic and Rare Diseases Information center internal data resources. The rare disease data enriched with additional biomedical data, including gene and phenotype ontologies, biological pathway data, and small molecule-target activity data, to create a knowledge graph (KG). Node embeddings were trained and clustered. We validated the disease clusters through semantic similarity and feature enrichment analysis.

RESULTS

Thirty-seven disease clusters were created with a mean size of 87 diseases. We validate the clusters quantitatively via semantic similarity based on the Orphanet Rare Disease Ontology. In addition, the clusters were analyzed for enrichment of associated genes, revealing that the enriched genes within clusters are highly related.

DISCUSSION

We demonstrate that node embeddings are an effective method for clustering diseases within a heterogenous KG. Semantically similar diseases and relevant enriched genes have been uncovered within the clusters. Connections between disease clusters and drugs are enumerated for follow-up efforts.

CONCLUSION

We lay out a method for clustering rare diseases using graph node embeddings. We develop an easy-to-maintain pipeline that can be updated when new data on rare diseases emerges. The embeddings themselves can be paired with other representation learning methods for other data types, such as drugs, to address other predictive modeling problems.

Multi-ontology embeddings approach on human-aligned multi-ontologies representation for gene-disease associations prediction

Objectives

Knowledge graphs and ontologies in the biomedical domain provide rich contextual knowledge for a variety of challenges. Employing that for knowledge-driven NLP tasks such as gene-disease association prediction represents a promising way to increase the predictive power of a model.

Methods

We investigated the power of infusing the embedding of two aligned ontologies as prior knowledge to the NLP models. We evaluated the performance of different models on some large-scale gene-disease association datasets and compared it with a model without incorporating contextualized knowledge (BERT).

Results

The experiments demonstrated that the knowledge-infused model slightly outperforms BERT by creating a small number of bridges. Thus, indicating that incorporating cross-references across ontologies can enhance the performance of base models without the need for more complex and costly training. However, further research is needed to explore the generalizability of the model. We expected that adding more bridges would bring further improvement based on the trend we observed in the experiments. In addition, the use of state-of-the-art knowledge graph embedding methods on a joint graph from connecting OGG and DOID with bridges also yielded promising results.

Conclusion

Our work shows that allowing language models to leverage structured knowledge from ontologies does come with clear advantages in the performance. Besides, the annotation stage brought out in this paper is constrained in reasonable complexity.

Hyperbolic hierarchical knowledge graph embeddings for biological entities

Predicting relationships between biological entities can greatly benefit important biomedical problems. Previous studies have attempted to represent biological entities and relationships in Euclidean space using embedding methods, which evaluate their semantic similarity by representing entities as numerical vectors. However, the limitation of these methods is that they cannot prevent the loss of latent hierarchical information when embedding large graph-structured data into Euclidean space, and therefore cannot capture the semantics of entities and relationships accurately. Hyperbolic spaces, such as Poincaré ball, are better suited for hierarchical modeling than Euclidean spaces. This is because hyperbolic spaces exhibit negative curvature, causing distances to grow exponentially as they approach the boundary. In this paper, we propose HEM, a hyperbolic hierarchical knowledge graph embedding model to generate vector representations of bio-entities. By encoding the entities and relations in the hyperbolic space, HEM can capture latent hierarchical information and improve the accuracy of biological entity representation. Notably, HEM can preserve rich information with a low dimension compared with the methods that encode entities in Euclidean space. Furthermore, we explore the performance of HEM in protein-protein interaction prediction and gene-disease association prediction tasks. Experimental results demonstrate the superior performance of HEM over state-of-the-art baselines. The data and code are available at : https://github.com/Nan-ll/HEM.

EGeRepDR: An enhanced genetic-based representation learning for drug repurposing using multiple biomedical sources

MOTIVATION

Drug repurposing (DR) is an imminent approach for identifying novel therapeutic indications for the available drugs and discovering novel drugs for previously untreatable diseases. Nowadays, DR has major attention in the pharmaceutical industry due to the high cost and time of launching new drugs to the market through traditional drug development. DR task majorly depends on genetic information since the drugs revert the modified Gene Expression (GE) of diseases to normal. Many of the existing studies have not considered the genetic importance of predicting the potential candidates.

METHOD

We proposed a novel multimodal framework that utilizes genetic aspects of drugs and diseases such as genes, pathways, gene signatures, or expression to enhance the performance of DR using various data sources. Firstly, the heterogeneous biological network (HBN) is constructed with three types of nodes namely drug, disease, and gene, and 4 types of edges similarities (drug, gene, and disease), drug-gene, gene-disease, and drug-disease. Next, a modified graph auto-encoder (GAE*) model is applied to learn the representation of drug and disease nodes using the topological structure and edge information. Secondly, the HBN is enhanced with the information extracted from biomedical literature and ontology using a novel semi-supervised pattern embedding-based bootstrapping model and novel DR perspective representation learning respectively to improve the prediction performance. Finally, our proposed system uses a neural network model to generate the probability score of drug-disease pairs.

RESULTS

We demonstrate the efficiency of the proposed model on various datasets and achieved outstanding performance in 5-fold cross-validation (AUC = 0.99, AUPR = 0.98). Further, we validated the top-ranked potential candidates using pathway analysis and proved that the known and predicted candidates share common genes in the pathways.

Prediction of Protein-Protein Interactions Using Vision Transformer and Language Model

The knowledge of protein-protein interaction (PPI) helps us to understand proteins' functions, the causes and growth of several diseases, and can aid in designing new drugs. The majority of existing PPI research has relied mainly on sequence-based approaches. With the availability of multi-omics datasets (sequence, 3D structure) and advancements in deep learning techniques, it is feasible to develop a deep multi-modal framework that fuses the features learned from different sources of information to predict PPI. In this work, we propose a multi-modal approach utilizing protein sequence and 3D structure. To extract features from the 3D structure of proteins, we use a pre-trained vision transformer model that has been fine-tuned on the structural representation of proteins. The protein sequence is encoded into a feature vector using a pre-trained language model. The feature vectors extracted from the two modalities are fused and then fed to the neural network classifier to predict the protein interactions. To showcase the effectiveness of the proposed methodology, we conduct experiments on two popular PPI datasets, namely, the human dataset and the S. cerevisiae dataset. Our approach outperforms the existing methodologies to predict PPI, including multi-modal approaches. We also evaluate the contributions of each modality by designing uni-modal baselines. We perform experiments with three modalities as well, having gene ontology as the third modality.

An Integrative Disease Information Network Approach to Similar Disease Detection

Disease similarity analysis impacts significantly in pathogenesis revealing, treatment recommending, and disease-causing genes predicting. Previous works study the disease similarity based on the semantics obtaining from biomedical ontologies (e.g., disease ontology) or the function of disease-causing molecules. However, such methods almost focus on a single perspective for obtaining disease features, which may lead to biased results for similar disease detection. To address this issue, we propose a disease information network-based integrative approach named MISSION for detecting similar diseases. By leveraging the associations between diseases and other biomedical entities, the disease information network is established first. Then, the disease similarity features extracted from the aspects of disease taxonomy, attributes, literature, and annotations are integrated into the disease information network. Finally, the top-k similar disease query is performed based on the integrative disease information. The experiments conducted on real-world datasets demonstrate that MISSION is effective and useful in similar disease detection.

Multi-domain knowledge graph embeddings for gene-disease association prediction

BACKGROUND

Predicting gene-disease associations typically requires exploring diverse sources of information as well as sophisticated computational approaches. Knowledge graph embeddings can help tackle these challenges by creating representations of genes and diseases based on the scientific knowledge described in ontologies, which can then be explored by machine learning algorithms. However, state-of-the-art knowledge graph embeddings are produced over a single ontology or multiple but disconnected ones, ignoring the impact that considering multiple interconnected domains can have on complex tasks such as gene-disease association prediction.

RESULTS

We propose a novel approach to predict gene-disease associations using rich semantic representations based on knowledge graph embeddings over multiple ontologies linked by logical definitions and compound ontology mappings. The experiments showed that considering richer knowledge graphs significantly improves gene-disease prediction and that different knowledge graph embeddings methods benefit more from distinct types of semantic richness.

CONCLUSIONS

This work demonstrated the potential for knowledge graph embeddings across multiple and interconnected biomedical ontologies to support gene-disease prediction. It also paved the way for considering other ontologies or tackling other tasks where multiple perspectives over the data can be beneficial. All software and data are freely available.

Enhanced disease-disease association with information enriched disease representation

OBJECTIVE

Quantification of disease-disease association (DDA) enables the understanding of disease relationships for discovering disease progression and finding comorbidity. For effective DDA strength calculation, there is a need to address the main challenge of integration of various biomedical aspects of DDA is to obtain an information rich disease representation.

MATERIALS AND METHODS

An enhanced and integrated DDA framework is developed that integrates enriched literature-based with concept-based DDA representation. The literature component of the proposed framework uses PubMed abstracts and consists of improved neural network model that classifies DDAs for an enhanced literature-based DDA representation. Similarly, an ontology-based joint multi-source association embedding model is proposed in the ontology component using Disease Ontology (DO), UMLS, claims insurance, clinical notes etc. Results and Discussion: The obtained information rich disease representation is evaluated on different aspects of DDA datasets such as Gene, Variant, Gene Ontology (GO) and a human rated benchmark dataset. The DDA scores calculated using the proposed method achieved a high correlation mainly in gene-based dataset. The quantified scores also shown better correlation of 0.821, when evaluated on human rated 213 disease pairs. In addition, the generated disease representation is proved to have substantial effect on correlation of DDA scores for different categories of disease pairs.

CONCLUSION

The enhanced context and semantic DDA framework provides an enriched disease representation, resulting in high correlated results with different DDA datasets. We have also presented the biological interpretation of disease pairs. The developed framework can also be used for deriving the strength of other biomedical associations.

Supporting SNOMED CT postcoordination with knowledge graph embeddings

SNOMED CT postcoordination is an underused mechanism that can help to implement advanced systems for the automatic extraction and encoding of clinical information from text. It allows defining non-existing SNOMED CT concepts by their relationships with existing ones. Manually building postcoordinated expressions is a difficult task. It requires a deep knowledge of the terminology and the support of specialized tools that barely exist. In order to support the building of postcoordinated expressions, we have implemented KGE4SCT: a method that suggests the corresponding SNOMED CT postcoordinated expression for a given clinical term. We leverage on the SNOMED CT ontology and its graph-like structure and use knowledge graph embeddings (KGEs). The objective of such embeddings is to represent in a vector space knowledge graph components (e.g. entities and relations) in a way that captures the structure of the graph. Then, we use vector similarity and analogies for obtaining the postcoordinated expression of a given clinical term. We obtained a semantic type accuracy of 98%, relationship accuracy of 90%, and analogy accuracy of 60%, with an overall completeness of postcoordination of 52% for the Spanish SNOMED CT version. We have also applied it to the English SNOMED CT version and outperformed state of the art methods in both, corpus generation for language model training for this task (improvement of 6% for analogy accuracy), and automatic postcoordination of SNOMED CT expressions, with an increase of 17% for partial conversion rate.

Clustering rare diseases within an ontology-enriched knowledge graph

Objective

Identifying sets of rare diseases with shared aspects of etiology and pathophysiology may enable drug repurposing and/or platform based therapeutic development. Toward that aim, we utilized an integrative knowledge graph-based approach to constructing clusters of rare diseases.

Materials and Methods

Data on 3,242 rare diseases were extracted from the National Center for Advancing Translational Science (NCATS) Genetic and Rare Diseases Information center (GARD) internal data resources. The rare disease data was enriched with additional biomedical data, including gene and phenotype ontologies, biological pathway data and small molecule-target activity data, to create a knowledge graph (KG). Node embeddings were used to convert nodes into vectors upon which k-means clustering was applied. We validated the disease clusters through semantic similarity and feature enrichment analysis.

Results

A node embedding model was trained on the ontology enriched rare disease KG and k-means clustering was applied to the embedding vectors resulting in 37 disease clusters with a mean size of 87 diseases. We validate the disease clusters quantitatively by looking at semantic similarity of clustered diseases, using the Orphanet Rare Disease Ontology. In addition, the clusters were analyzed for enrichment of associated genes, revealing that the enriched genes within clusters were shown to be highly related.

Discussion

We demonstrate that node embeddings are an effective method for clustering diseases within a heterogenous KG. Semantically similar diseases and relevant enriched genes have been uncovered within the clusters. Connections between disease clusters and approved or investigational drugs are enumerated for follow-up efforts.

Conclusion

Our study lays out a method for clustering rare diseases using the graph node embeddings. We develop an easy to maintain pipeline that can be updated when new data on rare diseases emerges. The embeddings themselves can be paired with other representation learning methods for other data types, such as drugs, to address other predictive modeling problems. Detailed subnetwork analysis and in-depth review of individual clusters may lead to translatable findings. Future work will focus on incorporation of additional data sources, with a particular focus on common disease data.

Building a search tool for compositely annotated entities using Transformer-based approach: Case study in Biosimulation Model Search Engine (BMSE)

The Transformer-based approaches to solving natural language processing (NLP) tasks such as BERT and GPT are gaining popularity due to their ability to achieve high performance. These approaches benefit from using enormous data sizes to create pre-trained models and the ability to understand the context of words in a sentence. Their use in the information retrieval domain is thought to increase effectiveness and efficiency. This paper demonstrates a BERT-based method (CASBERT) implementation to build a search tool over data annotated compositely using ontologies. The data was a collection of biosimulation models written using the CellML standard in the Physiome Model Repository (PMR). A biosimulation model structurally consists of basic entities of constants and variables that construct higher-level entities such as components, reactions, and the model. Finding these entities specific to their level is beneficial for various purposes regarding variable reuse, experiment setup, and model audit. Initially, we created embeddings representing compositely-annotated entities for constant and variable search (lowest level entity). Then, these low-level entity embeddings were vertically and efficiently combined to create higher-level entity embeddings to search components, models, images, and simulation setups. Our approach was general, so it can be used to create search tools with other data semantically annotated with ontologies - biosimulation models encoded in the SBML format, for example. Our tool is named Biosimulation Model Search Engine (BMSE).

Knowledge Graph Embeddings for ICU readmission prediction

BACKGROUND

Intensive Care Unit (ICU) readmissions represent both a health risk for patients,with increased mortality rates and overall health deterioration, and a financial burden for healthcare facilities. As healthcare became more data-driven with the introduction of Electronic Health Records (EHR), machine learning methods have been applied to predict ICU readmission risk. However, these methods disregard the meaning and relationships of data objects and work blindly over clinical data without taking into account scientific knowledge and context. Ontologies and Knowledge Graphs can help bridge this gap between data and scientific context, as they are computational artefacts that represent the entities of a domain and their relationships to each other in a formalized way.

METHODS AND RESULTS

We have developed an approach that enriches EHR data with semantic annotations to ontologies to build a Knowledge Graph. A patient's ICU stay is represented by Knowledge Graph embeddings in a contextualized manner, which are used by machine learning models to predict 30-days ICU readmissions. This approach is based on several contributions: (1) an enrichment of the MIMIC-III dataset with patient-oriented annotations to various biomedical ontologies; (2) a Knowledge Graph that defines patient data with biomedical ontologies; (3) a predictive model of ICU readmission risk that uses Knowledge Graph embeddings; (4) a variant of the predictive model that targets different time points during an ICU stay. Our predictive approaches outperformed both a baseline and state-of-the-art works achieving a mean Area Under the Receiver Operating Characteristic Curve of 0.827 and an Area Under the Precision-Recall Curve of 0.691. The application of this novel approach to help clinicians decide whether a patient can be discharged has the potential to prevent the readmission of [Formula: see text] of Intensive Care Unit patients, without unnecessarily prolonging the stay of those who would not require it.

CONCLUSION

The coupling of semantic annotation and Knowledge Graph embeddings affords two clear advantages: they consider scientific context and they are able to build representations of EHR information of different types in a common format. This work demonstrates the potential for impact that integrating ontologies and Knowledge Graphs into clinical machine learning applications can have.

MARPPI: boosting prediction of protein-protein interactions with multi-scale architecture residual network

Protein-protein interactions (PPIs) are a major component of the cellular biochemical reaction network. Rich sequence information and machine learning techniques reduce the dependence of exploring PPIs on wet experiments, which are costly and time-consuming. This paper proposes a PPI prediction model, multi-scale architecture residual network for PPIs (MARPPI), based on dual-channel and multi-feature. Multi-feature leverages Res2vec to obtain the association information between residues, and utilizes pseudo amino acid composition, autocorrelation descriptors and multivariate mutual information to achieve the amino acid composition and order information, physicochemical properties and information entropy, respectively. Dual channel utilizes multi-scale architecture improved ResNet network which extracts protein sequence features to reduce protein feature loss. Compared with other advanced methods, MARPPI achieves 96.03%, 99.01% and 91.80% accuracy in the intraspecific datasets of Saccharomyces cerevisiae, Human and Helicobacter pylori, respectively. The accuracy on the two interspecific datasets of Human-Bacillus anthracis and Human-Yersinia pestis is 97.29%, and 95.30%, respectively. In addition, results on specific datasets of disease (neurodegenerative and metabolic disorders) demonstrate the ability to detect hidden interactions. To better illustrate the performance of MARPPI, evaluations on independent datasets and PPIs network suggest that MARPPI can be used to predict cross-species interactions. The above shows that MARPPI can be regarded as a concise, efficient and accurate tool for PPI datasets.

TANGO: A GO-Term Embedding Based Method for Protein Semantic Similarity Prediction

We aim to quantitatively predict protein semantic similarities (PSS), which is vital to making biological discoveries. Previously, researchers commonly exploited Gene Ontology (GO) graphs (containing standardized hierarchically-organized GO terms for annotating distinct protein attributes) to learn GO term embeddings (vector representations) for quantifying protein attribute similarities and aggregate these embeddings to form protein embeddings for similarity measurement. However, two key properties of GO terms and annotated proteins are not yet well-explored by these learning-based methods: (1) taxonomy relations between GO terms; (2) GO terms' different contributions in describing protein semantics. In this paper, we propose TANGO, a new framework composed of a TAxoNomy-aware embedding module and an aggreGatiOn module. Our Embedding Module encodes taxonomic information into GO term embeddings by incorporating GO term topological distances in the GO graph hierarchy. Hence, distances between GO term embeddings can be used to more accurately measure shared meanings between correlated protein attributes. Our Aggregation Module automatically determines the contributions of GO terms when merging into the target protein embeddings, by mining GO term concept dependency relations in the GO graph and correlations in protein annotations. We conduct extensive experiments on several public datasets. On two PSS metrics, our new method significantly outperforms known methods by a large margin.

Analyzing Effect of Multi-Modality in Predicting Protein-Protein Interactions

Nowadays, multiple sources of information about proteins are available such as protein sequences, 3D structures, Gene Ontology (GO), etc. Most of the works on protein-protein interaction (PPI) identification had utilized these information about proteins, mainly sequence-based, but individually. The new advances in deep learning techniques allow us to leverage multiple sources/modalities of proteins, which complement each other. Some recent works have shown that multi-modal PPI models perform better than uni-modal approaches. This paper aims to investigate whether the performance of multi-modal PPI models is always consistent or depends on other factors such as dataset distribution, algorithms used to learn features, etc. We have used three modalities for this study: Protein sequence, 3D structure, and GO. Various techniques, including deep learning algorithms, are employed to extract features from multiple sources of proteins. These feature vectors from different modalities are then integrated in several combinations (bi-modal and tri-modal) to predict PPI. To conduct this study, we have used Human and S. cerevisiae PPI datasets. The obtained results demonstrate the potentiality of a multi-modal approach and deep learning techniques in predicting protein interactions. However, the predictive capability of a model for PPI depends on feature extraction methods as well. Also, increasing the modality does not always ensure performance improvement. In this study, the PPI model integrating two modalities outperforms the designed uni-modal and tri-modal PPI models.

mOWL: Python library for machine learning with biomedical ontologies

MOTIVATION

Ontologies contain formal and structured information about a domain and are widely used in bioinformatics for annotation and integration of data. Several methods use ontologies to provide background knowledge in machine learning tasks, which is of particular importance in bioinformatics. These methods rely on a set of common primitives that are not readily available in a software library; a library providing these primitives would facilitate the use of current machine learning methods with ontologies and the development of novel methods for other ontology-based biomedical applications.

RESULTS

We developed mOWL, a Python library for machine learning with ontologies formalized in the Web Ontology Language (OWL). mOWL implements ontology embedding methods that map information contained in formal knowledge bases and ontologies into vector spaces while preserving some of the properties and relations in ontologies, as well as methods to use these embeddings for similarity computation, deductive inference and zero-shot learning. We demonstrate mOWL on the knowledge-based prediction of protein-protein interactions using the gene ontology and gene-disease associations using phenotype ontologies.

AVAILABILITY AND IMPLEMENTATION

mOWL is freely available on https://github.com/bio-ontology-research-group/mowl and as a Python package in PyPi.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

A comprehensive update on CIDO: the community-based coronavirus infectious disease ontology

Background

The current COVID-19 pandemic and the previous SARS/MERS outbreaks of 2003 and 2012 have resulted in a series of major global public health crises. We argue that in the interest of developing effective and safe vaccines and drugs and to better understand coronaviruses and associated disease mechenisms it is necessary to integrate the large and exponentially growing body of heterogeneous coronavirus data. Ontologies play an important role in standard-based knowledge and data representation, integration, sharing, and analysis. Accordingly, we initiated the development of the community-based Coronavirus Infectious Disease Ontology (CIDO) in early 2020.

Results

As an Open Biomedical Ontology (OBO) library ontology, CIDO is open source and interoperable with other existing OBO ontologies. CIDO is aligned with the Basic Formal Ontology and Viral Infectious Disease Ontology. CIDO has imported terms from over 30 OBO ontologies. For example, CIDO imports all SARS-CoV-2 protein terms from the Protein Ontology, COVID-19-related phenotype terms from the Human Phenotype Ontology, and over 100 COVID-19 terms for vaccines (both authorized and in clinical trial) from the Vaccine Ontology. CIDO systematically represents variants of SARS-CoV-2 viruses and over 300 amino acid substitutions therein, along with over 300 diagnostic kits and methods. CIDO also describes hundreds of host-coronavirus protein-protein interactions (PPIs) and the drugs that target proteins in these PPIs. CIDO has been used to model COVID-19 related phenomena in areas such as epidemiology. The scope of CIDO was evaluated by visual analysis supported by a summarization network method. CIDO has been used in various applications such as term standardization, inference, natural language processing (NLP) and clinical data integration. We have applied the amino acid variant knowledge present in CIDO to analyze differences between SARS-CoV-2 Delta and Omicron variants. CIDO's integrative host-coronavirus PPIs and drug-target knowledge has also been used to support drug repurposing for COVID-19 treatment.

Conclusion

CIDO represents entities and relations in the domain of coronavirus diseases with a special focus on COVID-19. It supports shared knowledge representation, data and metadata standardization and integration, and has been used in a range of applications.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13326-022-00279-z.

Learning representations for gene ontology terms by jointly encoding graph structure and textual node descriptors

Measuring the semantic similarity between Gene Ontology (GO) terms is a fundamental step in numerous functional bioinformatics applications. To fully exploit the metadata of GO terms, word embedding-based methods have been proposed recently to map GO terms to low-dimensional feature vectors. However, these representation methods commonly overlook the key information hidden in the whole GO structure and the relationship between GO terms. In this paper, we propose a novel representation model for GO terms, named GT2Vec, which jointly considers the GO graph structure obtained by graph contrastive learning and the semantic description of GO terms based on BERT encoders. Our method is evaluated on a protein similarity task on a collection of benchmark datasets. The experimental results demonstrate the effectiveness of using a joint encoding graph structure and textual node descriptors to learn vector representations for GO terms.

How much do model organism phenotypes contribute to the computational identification of human disease genes?

Computing phenotypic similarity helps identify new disease genes and diagnose rare diseases. Genotype–phenotype data from orthologous genes in model organisms can compensate for lack of human data and increase genome coverage. In the past decade, cross-species phenotype comparisons have proven valuble, and several ontologies have been developed for this purpose. The relative contribution of different model organisms to computational identification of disease-associated genes is not fully explored. We used phenotype ontologies to semantically relate phenotypes resulting from loss-of-function mutations in model organisms to disease-associated phenotypes in humans. Semantic machine learning methods were used to measure the contribution of different model organisms to the identification of known human gene–disease associations. We found that mouse genotype–phenotype data provided the most important dataset in the identification of human disease genes by semantic similarity and machine learning over phenotype ontologies. Other model organisms' data did not improve identification over that obtained using the mouse alone, and therefore did not contribute significantly to this task. Our work impacts on the development of integrated phenotype ontologies, as well as for the use of model organism phenotypes in human genetic variant interpretation.

This article has an associated First Person interview with the first author of the paper.

Editor's choice: We investigated the use of model organism phenotypes in the computational identification of disease genes, identifying several data biases and concluding that mouse model phenotypes contribute most to computational disease gene identification.

Combining biomedical knowledge graphs and text to improve predictions for drug-target interactions and drug-indications

Biomedical knowledge is represented in structured databases and published in biomedical literature, and different computational approaches have been developed to exploit each type of information in predictive models. However, the information in structured databases and literature is often complementary. We developed a machine learning method that combines information from literature and databases to predict drug targets and indications. To effectively utilize information in published literature, we integrate knowledge graphs and published literature using named entity recognition and normalization before applying a machine learning model that utilizes the combination of graph and literature. We then use supervised machine learning to show the effects of combining features from biomedical knowledge and published literature on the prediction of drug targets and drug indications. We demonstrate that our approach using datasets for drug-target interactions and drug indications is scalable to large graphs and can be used to improve the ranking of targets and indications by exploiting features from either structure or unstructured information alone.

GOntoSim: a semantic similarity measure based on LCA and common descendants

The Gene Ontology (GO) is a controlled vocabulary that captures the semantics or context of an entity based on its functional role. Biomedical entities are frequently compared to each other to find similarities to help in data annotation and knowledge transfer. In this study, we propose GOntoSim, a novel method to determine the functional similarity between genes. GOntoSim quantifies the similarity between pairs of GO terms, by taking the graph structure and the information content of nodes into consideration. Our measure quantifies the similarity between the ancestors of the GO terms accurately. It also takes into account the common children of the GO terms. GOntoSim is evaluated using the entire Enzyme Dataset containing 10,890 proteins and 97,544 GO annotations. The enzymes are clustered and compared with the Gold Standard EC numbers. At level 1 of the EC Numbers for Molecular Function, GOntoSim achieves a purity score of 0.75 as compared to 0.47 and 0.51 GOGO and Wang. GOntoSim can handle the noisy IEA annotations. We achieve a purity score of 0.94 in contrast to 0.48 for both GOGO and Wang at level 1 of the EC Numbers with IEA annotations. GOntoSim can be freely accessed at (http://www.cbrlab.org/GOntoSim.html).

TransformerGO: predicting protein-protein interactions by modelling the attention between sets of gene ontology terms

MOTIVATION

Protein-protein interactions (PPIs) play a key role in diverse biological processes but only a small subset of the interactions has been experimentally identified. Additionally, high-throughput experimental techniques that detect PPIs are known to suffer various limitations, such as exaggerated false positives and negatives rates. The semantic similarity derived from the Gene Ontology (GO) annotation is regarded as one of the most powerful indicators for protein interactions. However, while computational approaches for prediction of PPIs have gained popularity in recent years, most methods fail to capture the specificity of GO terms.

RESULTS

We propose TransformerGO, a model that is capable of capturing the semantic similarity between GO sets dynamically using an attention mechanism. We generate dense graph embeddings for GO terms using an algorithmic framework for learning continuous representations of nodes in networks called node2vec. TransformerGO learns deep semantic relations between annotated terms and can distinguish between negative and positive interactions with high accuracy. TransformerGO outperforms classic semantic similarity measures on gold standard PPI datasets and state-of-the-art machine-learning-based approaches on large datasets from Saccharomyces cerevisiae and Homo sapiens. We show how the neural attention mechanism embedded in the transformer architecture detects relevant functional terms when predicting interactions.

AVAILABILITY AND IMPLEMENTATION

https://github.com/Ieremie/TransformerGO.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Biomedical data, computational methods and tools for evaluating disease-disease associations

In recent decades, exploring potential relationships between diseases has been an active research field. With the rapid accumulation of disease-related biomedical data, a lot of computational methods and tools/platforms have been developed to reveal intrinsic relationship between diseases, which can provide useful insights to the study of complex diseases, e.g. understanding molecular mechanisms of diseases and discovering new treatment of diseases. Human complex diseases involve both external phenotypic abnormalities and complex internal molecular mechanisms in organisms. Computational methods with different types of biomedical data from phenotype to genotype can evaluate disease-disease associations at different levels, providing a comprehensive perspective for understanding diseases. In this review, available biomedical data and databases for evaluating disease-disease associations are first summarized. Then, existing computational methods for disease-disease associations are reviewed and classified into five groups in terms of the usages of biomedical data, including disease semantic-based, phenotype-based, function-based, representation learning-based and text mining-based methods. Further, we summarize software tools/platforms for computation and analysis of disease-disease associations. Finally, we give a discussion and summary on the research of disease-disease associations. This review provides a systematic overview for current disease association research, which could promote the development and applications of computational methods and tools/platforms for disease-disease associations.

Anc2vec: embedding gene ontology terms by preserving ancestors relationships

The gene ontology (GO) provides a hierarchical structure with a controlled vocabulary composed of terms describing functions and localization of gene products. Recent works propose vector representations, also known as embeddings, of GO terms that capture meaningful information about them. Significant performance improvements have been observed when these representations are used on diverse downstream tasks, such as the measurement of semantic similarity between GO terms and functional similarity between proteins. Despite the success shown by these approaches, existing embeddings of GO terms still fail to capture crucial structural features of the GO. Here, we present anc2vec, a novel protocol based on neural networks for constructing vector representations of GO terms by preserving three important ontological features: its ontological uniqueness, ancestors hierarchy and sub-ontology membership. The advantages of using anc2vec are demonstrated by systematic experiments on diverse tasks: visualization, sub-ontology prediction, inference of structurally related terms, retrieval of terms from aggregated embeddings, and prediction of protein-protein interactions. In these tasks, experimental results show that the performance of anc2vec representations is better than those of recent approaches. This demonstrates that higher performances on diverse tasks can be achieved by embeddings when the structure of the GO is better represented. Full source code and data are available at https://github.com/sinc-lab/anc2vec.

Evaluating semantic similarity methods for comparison of text-derived phenotype profiles

Background

Semantic similarity is a valuable tool for analysis in biomedicine. When applied to phenotype profiles derived from clinical text, they have the capacity to enable and enhance ‘patient-like me’ analyses, automated coding, differential diagnosis, and outcome prediction. While a large body of work exists exploring the use of semantic similarity for multiple tasks, including protein interaction prediction, and rare disease differential diagnosis, there is less work exploring comparison of patient phenotype profiles for clinical tasks. Moreover, there are no experimental explorations of optimal parameters or better methods in the area.

Methods

We develop a platform for reproducible benchmarking and comparison of experimental conditions for patient phentoype similarity. Using the platform, we evaluate the task of ranking shared primary diagnosis from uncurated phenotype profiles derived from all text narrative associated with admissions in the medical information mart for intensive care (MIMIC-III).

Results

300 semantic similarity configurations were evaluated, as well as one embedding-based approach. On average, measures that did not make use of an external information content measure performed slightly better, however the best-performing configurations when measured by area under receiver operating characteristic curve and Top Ten Accuracy used term-specificity and annotation-frequency measures.

Conclusion

We identified and interpreted the performance of a large number of semantic similarity configurations for the task of classifying diagnosis from text-derived phenotype profiles in one setting. We also provided a basis for further research on other settings and related tasks in the area.

DeepSVP: integration of genotype and phenotype for structural variant prioritization using deep learning

MOTIVATION

Structural genomic variants account for much of human variability and are involved in several diseases. Structural variants are complex and may affect coding regions of multiple genes, or affect the functions of genomic regions in different ways from single nucleotide variants. Interpreting the phenotypic consequences of structural variants relies on information about gene functions, haploinsufficiency or triplosensitivity and other genomic features. Phenotype-based methods to identifying variants that are involved in genetic diseases combine molecular features with prior knowledge about the phenotypic consequences of altering gene functions. While phenotype-based methods have been applied successfully to single nucleotide variants as well as short insertions and deletions, the complexity of structural variants makes it more challenging to link them to phenotypes. Furthermore, structural variants can affect a large number of coding regions, and phenotype information may not be available for all of them.

RESULTS

We developed DeepSVP, a computational method to prioritize structural variants involved in genetic diseases by combining genomic and gene functions information. We incorporate phenotypes linked to genes, functions of gene products, gene expression in individual cell types and anatomical sites of expression, and systematically relate them to their phenotypic consequences through ontologies and machine learning. DeepSVP significantly improves the success rate of finding causative variants in several benchmarks and can identify novel pathogenic structural variants in consanguineous families.

AVAILABILITY AND IMPLEMENTATION

https://github.com/bio-ontology-research-group/DeepSVP.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

GoVec: Gene Ontology Representation Learning Using Weighted Heterogeneous Graph and Meta-Path

Biomedical knowledge graphs are crucial to support data-intensive applications in the life sciences and health care. These graphs can be extended by generating a heterogeneous graph that contains both ontology terms and biomedical entities. However, state-of-the-art approaches for Gene Ontology representation learnings are constrained to homogeneous graphs that cannot represent different node types and relations. To address this limitation, we present GoVec to produce representations seamlessly for both ontologies and biological entities by utilizing meta-path-based representation learning in the heterogeneous graph. The resulting vectors can be used in many bioinformatics applications, particularly for calculating semantic similarity and extracting relations among biological entities. We verify the approach's usefulness by comparing the resulting semantic similarities with the manually produced similarities by the experts. Furthermore, the superiority of the GoVec is shown by an extensive set of quantitative and qualitative evaluations. Two downstream tasks, including protein-protein interaction and protein family similarity, are evaluated in comparison with many state-of-the-art approaches. Finally, as a qualitative visual representation, the separability of various protein families is examined and visually separable groups of proteins are generated, which shows the capability of GoVec representations to embed functional semantics into the vectors.

Network-based protein-protein interaction prediction method maps perturbations of cancer interactome

The perturbations of protein-protein interactions (PPIs) were found to be the main cause of cancer. Previous PPI prediction methods which were trained with non-disease general PPI data were not compatible to map the PPI network in cancer. Therefore, we established a novel cancer specific PPI prediction method dubbed NECARE, which was based on relational graph convolutional network (R-GCN) with knowledge-based features. It achieved the best performance with a Matthews correlation coefficient (MCC) = 0.84±0.03 and an F1 = 91±2% compared with other methods. With NECARE, we mapped the cancer interactome atlas and revealed that the perturbations of PPIs were enriched on 1362 genes, which were named cancer hub genes. Those genes were found to over-represent with mutations occurring at protein-macromolecules binding interfaces. Furthermore, over 56% of cancer treatment-related genes belonged to hub genes and they were significantly related to the prognosis of 32 types of cancers. Finally, by coimmunoprecipitation, we confirmed that the NECARE prediction method was highly reliable with a 90% accuracy. Overall, we provided the novel network-based cancer protein-protein interaction prediction method and mapped the perturbation of cancer interactome. NECARE is available at: https://github.com/JiajunQiu/NECARE.

Dimensional reduction of phenotypes from 53 000 mouse models reveals a diverse landscape of gene function

Animal models have long been used to study gene function and the impact of genetic mutations on phenotype. Through the research efforts of thousands of research groups, systematic curation of published literature and high-throughput phenotyping screens, the collective body of knowledge for the mouse now covers the majority of protein-coding genes. We here collected data for over 53 000 mouse models with mutations in over 15 000 genomic markers and characterized by more than 254 000 annotations using more than 9000 distinct ontology terms. We investigated dimensional reduction and embedding techniques as means to facilitate access to this diverse and high-dimensional information. Our analyses provide the first visual maps of the landscape of mouse phenotypic diversity. We also summarize some of the difficulties in producing and interpreting embeddings of sparse phenotypic data. In particular, we show that data preprocessing, filtering and encoding have as much impact on the final embeddings as the process of dimensional reduction. Nonetheless, techniques developed in the context of dimensional reduction create opportunities for explorative analysis of this large pool of public data, including for searching for mouse models suited to study human diseases.

AVAILABILITY AND IMPLEMENTATION

Source code for analysis scripts is available on GitHub at https://github.com/tkonopka/mouse-embeddings. The data underlying this article are available in Zenodo at https://doi.org/10.5281/zenodo.4916171.

CONTACT

t.konopka@qmul.ac.uk.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics Advances online.

HiG2Vec: hierarchical representations of Gene Ontology and genes in the Poincaré ball

MOTIVATION

Knowledge manipulation of Gene Ontology (GO) and Gene Ontology Annotation (GOA) can be done primarily by using vector representation of GO terms and genes. Previous studies have represented GO terms and genes or gene products in Euclidean space to measure their semantic similarity using an embedding method such as the Word2Vec-based method to represent entities as numeric vectors. However, this method has the limitation that embedding large graph-structured data in the Euclidean space cannot prevent a loss of information of latent hierarchies, thus precluding the semantics of GO and GOA from being captured optimally. On the other hand, hyperbolic spaces such as the Poincaré balls are more suitable for modeling hierarchies, as they have a geometric property in which the distance increases exponentially as it nears the boundary because of negative curvature.

RESULTS

In this article, we propose hierarchical representations of GO and genes (HiG2Vec) by applying Poincaré embedding specialized in the representation of hierarchy through a two-step procedure: GO embedding and gene embedding. Through experiments, we show that our model represents the hierarchical structure better than other approaches and predicts the interaction of genes or gene products similar to or better than previous studies. The results indicate that HiG2Vec is superior to other methods in capturing the GO and gene semantics and in data utilization as well. It can be robustly applied to manipulate various biological knowledge.

AVAILABILITYAND IMPLEMENTATION

https://github.com/JaesikKim/HiG2Vec.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

BioERP: biomedical heterogeneous network-based self-supervised representation learning approach for entity relationship predictions

MOTIVATION

Predicting entity relationship can greatly benefit important biomedical problems. Recently, a large amount of biomedical heterogeneous networks (BioHNs) are generated and offer opportunities for developing network-based learning approaches to predict relationships among entities. However, current researches slightly explored BioHNs-based self-supervised representation learning methods, and are hard to simultaneously capturing local- and global-level association information among entities.

RESULTS

In this study, we propose a biomedical heterogeneous network-based self-supervised representation learning approach for entity relationship predictions, termed BioERP. A self-supervised meta path detection mechanism is proposed to train a deep Transformer encoder model that can capture the global structure and semantic feature in BioHNs. Meanwhile, a biomedical entity mask learning strategy is designed to reflect local associations of vertices. Finally, the representations from different task models are concatenated to generate two-level representation vectors for predicting relationships among entities. The results on eight datasets show BioERP outperforms 30 state-of-the-art methods. In particular, BioERP reveals great performance with results close to 1 in terms of AUC and AUPR on the drug-target interaction predictions. In summary, BioERP is a promising bio-entity relationship prediction approach.

AVAILABILITY

Source code and data can be downloaded from https://github.com/pengsl-lab/BioERP.git.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Semantic similarity and machine learning with ontologies

Ontologies have long been employed in the life sciences to formally represent and reason over domain knowledge and they are employed in almost every major biological database. Recently, ontologies are increasingly being used to provide background knowledge in similarity-based analysis and machine learning models. The methods employed to combine ontologies and machine learning are still novel and actively being developed. We provide an overview over the methods that use ontologies to compute similarity and incorporate them in machine learning methods; in particular, we outline how semantic similarity measures and ontology embeddings can exploit the background knowledge in ontologies and how ontologies can provide constraints that improve machine learning models. The methods and experiments we describe are available as a set of executable notebooks, and we also provide a set of slides and additional resources at https://github.com/bio-ontology-research-group/machine-learning-with-ontologies.

OWL2Vec*: embedding of OWL ontologies

Semantic embedding of knowledge graphs has been widely studied and used for prediction and statistical analysis tasks across various domains such as Natural Language Processing and the Semantic Web. However, less attention has been paid to developing robust methods for embedding OWL (Web Ontology Language) ontologies, which contain richer semantic information than plain knowledge graphs, and have been widely adopted in domains such as bioinformatics. In this paper, we propose a random walk and word embedding based ontology embedding method named , which encodes the semantics of an OWL ontology by taking into account its graph structure, lexical information and logical constructors. Our empirical evaluation with three real world datasets suggests that benefits from these three different aspects of an ontology in class membership prediction and class subsumption prediction tasks. Furthermore, often significantly outperforms the state-of-the-art methods in our experiments.

A representation model for biological entities by fusing structured axioms with unstructured texts

MOTIVATION

Structured semantic resources, for example, biological knowledge bases and ontologies, formally define biological concepts, entities and their semantic relationships, manifested as structured axioms and unstructured texts (e.g. textual definitions). The resources contain accurate expressions of biological reality and have been used by machine-learning models to assist intelligent applications like knowledge discovery. The current methods use both the axioms and definitions as plain texts in representation learning (RL). However, since the axioms are machine-readable while the natural language is human-understandable, difference in meaning of token and structure impedes the representations to encode desirable biological knowledge.

RESULTS

We propose ERBK, a RL model of bio-entities. Instead of using the axioms and definitions as a textual corpus, our method uses knowledge graph embedding method and deep convolutional neural models to encode the axioms and definitions respectively. The representations could not only encode more underlying biological knowledge but also be further applied to zero-shot circumstance where existing approaches fall short. Experimental evaluations show that ERBK outperforms the existing methods for predicting protein-protein interactions and gene-disease associations. Moreover, it shows that ERBK still maintains promising performance under the zero-shot circumstance. We believe the representations and the method have certain generality and could extend to other types of bio-relation.

AVAILABILITY AND IMPLEMENTATION

The source code is available at the gitlab repository https://gitlab.com/BioAI/erbk.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Predicting candidate genes from phenotypes, functions and anatomical site of expression

Motivation

Over the past years, many computational methods have been developed to incorporate information about phenotypes for disease–gene prioritization task. These methods generally compute the similarity between a patient’s phenotypes and a database of gene-phenotype to find the most phenotypically similar match. The main limitation in these methods is their reliance on knowledge about phenotypes associated with particular genes, which is not complete in humans as well as in many model organisms, such as the mouse and fish. Information about functions of gene products and anatomical site of gene expression is available for more genes and can also be related to phenotypes through ontologies and machine-learning models.

Results

We developed a novel graph-based machine-learning method for biomedical ontologies, which is able to exploit axioms in ontologies and other graph-structured data. Using our machine-learning method, we embed genes based on their associated phenotypes, functions of the gene products and anatomical location of gene expression. We then develop a machine-learning model to predict gene–disease associations based on the associations between genes and multiple biomedical ontologies, and this model significantly improves over state-of-the-art methods. Furthermore, we extend phenotype-based gene prioritization methods significantly to all genes, which are associated with phenotypes, functions or site of expression.

Availability and implementation

Software and data are available at https://github.com/bio-ontology-research-group/DL2Vec.

Supplementary information

Supplementary data are available at Bioinformatics online.

Towards similarity-based differential diagnostics for common diseases

Ontology-based phenotype profiles have been utilised for the purpose of differential diagnosis of rare genetic diseases, and for decision support in specific disease domains. Particularly, semantic similarity facilitates diagnostic hypothesis generation through comparison with disease phenotype profiles. However, the approach has not been applied for differential diagnosis of common diseases, or generalised clinical diagnostics from uncurated text-derived phenotypes. In this work, we describe the development of an approach for deriving patient phenotype profiles from clinical narrative text, and apply this to text associated with MIMIC-III patient visits. We then explore the use of semantic similarity with those text-derived phenotypes to classify primary patient diagnosis, comparing the use of patient-patient similarity and patient-disease similarity using phenotype-disease profiles previously mined from literature. We also consider a combined approach, in which literature-derived phenotypes are extended with the content of text-derived phenotypes we mined from 500 patients. The results reveal a powerful approach, showing that in one setting, uncurated text phenotypes can be used for differential diagnosis of common diseases, making use of information both inside and outside the setting. While the methods themselves should be explored for further optimisation, they could be applied to a variety of clinical tasks, such as differential diagnosis, cohort discovery, document and text classification, and outcome prediction.

nhKcr: a new bioinformatics tool for predicting crotonylation sites on human nonhistone proteins based on deep learning

Lysine crotonylation (Kcr) is a newly discovered type of protein post-translational modification and has been reported to be involved in various pathophysiological processes. High-resolution mass spectrometry is the primary approach for identification of Kcr sites. However, experimental approaches for identifying Kcr sites are often time-consuming and expensive when compared with computational approaches. To date, several predictors for Kcr site prediction have been developed, most of which are capable of predicting crotonylation sites on either histones alone or mixed histone and nonhistone proteins together. These methods exhibit high diversity in their algorithms, encoding schemes, feature selection techniques and performance assessment strategies. However, none of them were designed for predicting Kcr sites on nonhistone proteins. Therefore, it is desirable to develop an effective predictor for identifying Kcr sites from the large amount of nonhistone sequence data. For this purpose, we first provide a comprehensive review on six methods for predicting crotonylation sites. Second, we develop a novel deep learning-based computational framework termed as CNNrgb for Kcr site prediction on nonhistone proteins by integrating different types of features. We benchmark its performance against multiple commonly used machine learning classifiers (including random forest, logitboost, naïve Bayes and logistic regression) by performing both 10-fold cross-validation and independent test. The results show that the proposed CNNrgb framework achieves the best performance with high computational efficiency on large datasets. Moreover, to facilitate users' efforts to investigate Kcr sites on human nonhistone proteins, we implement an online server called nhKcr and compare it with other existing tools to illustrate the utility and robustness of our method. The nhKcr web server and all the datasets utilized in this study are freely accessible at http://nhKcr.erc.monash.edu/.

Evaluating Disease Similarity Based on Gene Network Reconstruction and Representation

MOTIVATION

Quantifying the associations between diseases is of great significance in increasing our understanding of disease biology, improving disease diagnosis, re-positioning, and developing drugs. Therefore, in recent years, the research of disease similarity has received a lot of attention in the field of bioinformatics. Previous work has shown that the combination of the ontology (such as disease ontology and gene ontology) and disease-gene interactions are worthy to be regarded to elucidate diseases and disease associations. However, most of them are either based on the overlap between disease-related gene sets or distance within the ontology's hierarchy. The diseases in these methods are represented by discrete or sparse feature vectors, which cannot grasp the deep semantic information of diseases. Recently, deep representation learning has been widely studied and gradually applied to various fields of bioinformatics. Based on the hypothesis that disease representation depends on its related gene representations, we propose a disease representation model using two most representative gene resources HumanNet and Gene Ontology to construct a new gene network and learn gene (disease) representations. The similarity between two diseases is computed by the cosine similarity of their corresponding representations.

RESULTS

We propose a novel approach to compute disease similarity, which integrates two important factors disease-related genes and gene ontology hierarchy to learn disease representation based on deep representation learning. Under the same experimental settings, the AUC value of our method is 0.8074, which improves the most competitive baseline method by 10.1%. The quantitative and qualitative experimental results show that our model can learn effective disease representations and improve the accuracy of disease similarity computation significantly.

AVAILABILITY

The research shows that this method has certain applicability in the prediction of gene-related diseases, the migration of disease treatment methods, drug development, and so on.

SUPPLEMENTARY INFORMATION

Supplementary data are available at https://github.com/catly/disease_similarity.

Graph embeddings on gene ontology annotations for protein-protein interaction prediction

Background

Protein–protein interaction (PPI) prediction is an important task towards the understanding of many bioinformatics functions and applications, such as predicting protein functions, gene-disease associations and disease-drug associations. However, many previous PPI prediction researches do not consider missing and spurious interactions inherent in PPI networks. To address these two issues, we define two corresponding tasks, namely missing PPI prediction and spurious PPI prediction, and propose a method that employs graph embeddings that learn vector representations from constructed Gene Ontology Annotation (GOA) graphs and then use embedded vectors to achieve the two tasks. Our method leverages on information from both term–term relations among GO terms and term-protein annotations between GO terms and proteins, and preserves properties of both local and global structural information of the GO annotation graph.

Results

We compare our method with those methods that are based on information content (IC) and one method that is based on word embeddings, with experiments on three PPI datasets from STRING database. Experimental results demonstrate that our method is more effective than those compared methods.

Conclusion

Our experimental results demonstrate the effectiveness of using graph embeddings to learn vector representations from undirected GOA graphs for our defined missing and spurious PPI tasks.

A semantic similarity based methodology for predicting protein-protein interactions: Evaluation with P53-interacting kinases

Biomedical literature contains unstructured, rich information regarding proteins, ligands, diseases as well as biological pathways in which they are involved. Systematically analyzing such textual corpus has the potential for biomedical discovery of new protein-protein interactions and hidden drug indications. For this purpose, we have investigated a methodology that is based on a well-established text mining tool, Word2Vec, for the analysis of PubMed full text articles to derive word embeddings, and the use of a simple semantic similarity comparison either by itself or in conjunction with k-Nearest Neighbor (kNN) technique for the prediction of new relationships. To test this methodology, three lines of retrospective analyses of a dataset with known P53-interacting proteins have been conducted. First, we demonstrated that Word2Vec semantic similarity can infer functional relatedness among all kinases known to interact with P53. Second, in a series of time-split experiments, we demonstrated that both a simple similarity comparison and kNN models built with papers published up to a certain year were able to discover P53 interactors described in later publications. Third, in a different scenario of time-split experiments, we examined the predictions of P53-interacting proteins based on the kNN models built on data prior to a certain split year for different time ranges past that year, and found that the cumulative number of correct predictions was indeed increasing with time. We conclude that text mining of research papers in the PubMed literature based on Word2Vec analysis followed by a simple similarity comparison or kNN modeling affords excellent predictions of protein-protein interactions between P53 and kinases, and should have wide applications in translational biomedical studies such as repurposing of existing drugs, drug-drug interaction, and elucidation of mechanisms of action for drugs.

Formal axioms in biomedical ontologies improve analysis and interpretation of associated data

Motivation

Over the past years, significant resources have been invested into formalizing biomedical ontologies. Formal axioms in ontologies have been developed and used to detect and ensure ontology consistency, find unsatisfiable classes, improve interoperability, guide ontology extension through the application of axiom-based design patterns and encode domain background knowledge. The domain knowledge of biomedical ontologies may have also the potential to provide background knowledge for machine learning and predictive modelling.

Results

We use ontology-based machine learning methods to evaluate the contribution of formal axioms and ontology meta-data to the prediction of protein–protein interactions and gene–disease associations. We find that the background knowledge provided by the Gene Ontology and other ontologies significantly improves the performance of ontology-based prediction models through provision of domain-specific background knowledge. Furthermore, we find that the labels, synonyms and definitions in ontologies can also provide background knowledge that may be exploited for prediction. The axioms and meta-data of different ontologies contribute to improving data analysis in a context-specific manner. Our results have implications on the further development of formal knowledge bases and ontologies in the life sciences, in particular as machine learning methods are more frequently being applied. Our findings motivate the need for further development, and the systematic, application-driven evaluation and improvement, of formal axioms in ontologies.

Availability and implementation

https://github.com/bio-ontology-research-group/tsoe.

Supplementary information

Supplementary data are available at Bioinformatics online.

Conjoint Feature Representation of GO and Protein Sequence for PPI Prediction Based on an Inception RNN Attention Network

Protein-protein interactions (PPIs) are pivotal for cellular functions and biological processes. In the past years, computational methods using amino acid sequences and gene ontology (GO) annotations of proteins for prioritizing PPIs have provided important references for biological experiments in the wet lab. Despite the current success, sequence information and ontological annotation in semantic representation have not been integrated into current methods. We propose a deep-learning-based PPI prediction methodology conjointly featuring sequence information and GO annotation. First, we adopt a word-embedding tool, the NCBI-blueBERT model pre-trained on PubMed, to map the GO terms into their semantic vectors. Then, the GO semantic vectors and protein sequence vector serve as the input of the proposed inception recurrent neural network (RNN) attention network (IRAN). The IRAN captures the spatial relationship and the potential sequential feature of the protein sequence and ontological annotation semantics. The extensive experimental results on 12 benchmarks demonstrate that our method achieves superiority over state-of-the-art baselines. In the yeast dataset of a binary PPI prediction, our method improved the performance with the Matthews correlation coefficient increasing from 94.2% to 98.2% and the accuracy from 97.1% to 98.2%. The analogous results were also obtained in other comparison evaluations.

Knowledge-Based Biomedical Data Science

Knowledge-based biomedical data science involves the design and implementation of computer systems that act as if they knew about biomedicine. Such systems depend on formally represented knowledge in computer systems, often in the form of knowledge graphs. Here we survey recent progress in systems that use formally represented knowledge to address data science problems in both clinical and biological domains, as well as progress on approaches for creating knowledge graphs. Major themes include the relationships between knowledge graphs and machine learning, the use of natural language processing to construct knowledge graphs, and the expansion of novel knowledge-based approaches to clinical and biological domains.