Pharmacovigilance and clinical decision support systems utilize structured drug safety data to guide medical practice. However, existing datasets frequently depend on terminologies such as MedDRA, which limits their semantic reasoning capabilities and their interoperability with Semantic Web ontologies and knowledge graphs. To address this gap, we developed SIDEKICK, a knowledge graph that standardizes drug indications, contraindications, and adverse reactions from FDA Structured Product Labels. We developed and used a workflow based on Large Language Model (LLM) extraction and Graph-Retrieval Augmented Generation (Graph RAG) for ontology mapping. We processed over 50,000 drug labels and mapped terms to the Human Phenotype Ontology (HPO), the MONDO Disease Ontology, and RxNorm. Our semantically integrated resource outperforms the SIDER and ONSIDES databases when applied to the task of drug repurposing by side effect similarity. We serialized the dataset as a Resource Description Framework (RDF) graph and employed the Semanticscience Integrated Ontology (SIO) as upper level ontology to further improve interoperability. Consequently, SIDEKICK enables automated safety surveillance and phenotype-based similarity analysis for drug repurposing.

MOTIVATION: Knowledge graphs and structural causal models have each proven valuable for organizing biomedical knowledge and estimating causal effects, but remain largely disconnected: knowledge graphs encode qualitative relationships focusing on facts and deductive reasoning without formal probabilistic semantics, while causal models lack integration with background knowledge in knowledge graphs and have no access to the deductive reasoning capabilities that knowledge graphs provide. RESULTS: To bridge this gap, we introduce a novel formulation of Causal Knowledge Graphs (CKGs) which extend knowledge graphs with formal causal semantics, preserving their deductive capabilities while enabling principled causal inference. CKGs support deconfounding via explicitly marked causal edges and facilitate hypothesis formulation aligned with both encoded and entailed background knowledge. We constructed a Drug-Disease CKG (DD-CKG) integrating disease progression pathways, drug indications, side-effects, and hierarchical disease classification to enable automated large-scale mediation analysis. Applied to UK Biobank and MIMIC-IV cohorts, we tested whether drugs mediate effects between indications and downstream disease progression, adjusting for confounders inferred from the DD-CKG. Our approach successfully reproduced known adverse drug reactions with high precision while identifying previously undocumented significant candidate adverse effects. Further validation through side effect similarity analysis demonstrated that combining our predicted drug effects with established databases significantly improves the prediction of shared drug indications, supporting the clinical relevance of our novel findings. These results demonstrate that our methodology provides a generalizable, knowledge-driven framework for scalable causal inference. AVAILABILITY AND IMPLEMENTATION: The data is available through https://github.com/bio-ontology-research-group/Mediation-Analysis-using-Causal-Knowledge-Graph.

MOTIVATION: Predicting gene-disease associations (GDAs) can be framed as a ranking problem where genes are ranked for a query disease based on features such as phenotypic similarity. By describing phenotypes using phenotype ontologies, ontology-based semantic similarity measures can be used. However, traditional semantic similarity measures use only the ontology taxonomy. Recent methods based on ontology embeddings compare phenotypes in latent space; these methods can use all ontology axioms as well as a supervised signal, but are inherently transductive, i.e. query entities must already be known at the time of learning embeddings, and therefore these methods do not generalize to novel diseases (sets of phenotypes) at inference time. RESULTS: We developed INDIGENA, an inductive disease-gene association method for ranking genes based on a set of phenotypes. Our method first uses a graph projection to map axioms from phenotype ontologies to a graph structure, and then uses graph embeddings to create latent representations of phenotypes. We use an explicit aggregation strategy to combine phenotype embeddings into representations of genes or diseases, allowing us to generalize to novel sets of phenotypes. We also develop a method to make the phenotype embeddings and the similarity measure task-specific by including a supervised signal from known GDAs. We apply our method to mouse models of human disease and demonstrate that we can significantly improve over the inductive semantic similarity baseline measures, and reach a performance similar to transductive methods for predicting GDAs while being more general. AVAILABILITY AND IMPLEMENTATION: https://github.com/bio-ontology-research-group/indigena.

Ontologies are widely used for representing domain knowledge and meta data, playing an increasingly important role in Information Systems, the Semantic Web, Bioinformatics and many other domains. However, logical reasoning that ontologies can directly support are quite limited in learning, approximation and prediction. One straightforward solution is to integrate statistical analysis and machine learning. To this end, automatically learning vector representation for knowledge of an ontology i.e.,ontology embeddinghas been widely investigated. Numerous papers have been published on ontology embedding, but a lack of systematic reviews hinders researchers from gaining a comprehensive understanding of this field. To bridge this gap, we write this survey paper, which first introduces different kinds of semantics of ontologies and formally defines ontology embedding as well as its property of faithfulness. Based on this, it systematically categorizes and analyses a relatively complete set of over 80 papers, according to the ontologies they aim at and their technical solutions including geometric modeling, sequence modeling and graph propagation. This survey also introduces the applications of ontology embedding in ontology engineering, machine learning augmentation and life sciences, presents a new library mOWL and discusses the challenges and future directions.

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.

MOTIVATION: Identifying causal relations between diseases allows for the study of shared pathways, biological mechanisms, and inter-disease risks. Such causal relations can facilitate the identification of potential disease precursors and candidates for drug re-purposing. However, computational methods often lack access to these causal relations. Few approaches have been developed to automatically extract causal relationships between diseases from unstructured text, but they are often only focused on a small number of diseases, lack validation of the extracted causal relations, or do not make their data available. RESULTS: We automatically mined statements asserting a causal relation between diseases from the scientific literature by leveraging lexical patterns. Following automated mining of causal relations, we mapped the diseases to the International Classification of Diseases (ICD) identifiers to allow the direct application to clinical data. We provide quantitative and qualitative measures to evaluate the mined causal relations and compare to UK Biobank diagnosis data as a completely independent data source. The validated causal associations were used to create a directed acyclic graph that can be used by causal inference frameworks. We demonstrate the utility of our causal network by performing causal inference using the do-calculus, using relations within the graph to construct and improve polygenic risk scores, and disentangle the pleiotropic effects of variants. AVAILABILITY AND IMPLEMENTATION: The data are available through https://github.com/bio-ontology-research-group/causal-relations-between-diseases.

Automated protein function prediction is a crucial and widely studied problem in bioinformatics. Computationally, protein function is a multilabel classification problem where only positive samples are defined and there is a large number of unlabeled annotations. Most existing methods rely on the assumption that the unlabeled set of protein function annotations are negatives, inducing the false negative issue, where potential positive samples are trained as negatives. We introduce a novel approach named PU-GO, wherein we address function prediction as a positive-unlabeled ranking problem. We apply empirical risk minimization, i.e. we minimize the classification risk of a classifier where class priors are obtained from the Gene Ontology hierarchical structure. We show that our approach is more robust than other state-of-the-art methods on similarity-based and time-based benchmark datasets. AVAILABILITY AND IMPLEMENTATION: Data and code are available at https://github.com/bio-ontology-research-group/PU-GO.

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.

BACKGROUND: Identifying variants associated with diseases is a challenging task in medical genetics research. Current studies that prioritize variants within individual genomes generally rely on known variants, evidence from literature and genomes, and patient symptoms and clinical signs. The functionalities of the existing tools, which rank variants based on given patient symptoms and clinical signs, are restricted to the coverage of ontologies such as the Human Phenotype Ontology (HPO). However, most clinicians do not limit themselves to HPO while describing patient symptoms/signs and their associated variants/genes. There is thus a need for an automated tool that can prioritize variants based on freely expressed patient symptoms and clinical signs. RESULTS: STARVar is a Symptom-based Tool for Automatic Ranking of Variants using evidence from literature and genomes. STARVar uses patient symptoms and clinical signs, either linked to HPO or expressed in free text format. It returns a ranked list of variants based on a combined score from two classifiers utilizing evidence from genomics and literature. STARVar improves over related tools on a set of synthetic patients. In addition, we demonstrated its distinct contribution to the domain on another synthetic dataset covering publicly available clinical genotype-phenotype associations by using symptoms and clinical signs expressed in free text format. CONCLUSIONS: STARVar stands as a unique and efficient tool that has the advantage of ranking variants with flexibly expressed patient symptoms in free-form text. Therefore, STARVar can be easily integrated into bioinformatics workflows designed to analyze disease-associated genomes. AVAILABILITY: STARVar is freely available from https://github.com/bio-ontology-research-group/STARVar .

Most real-world knowledge graphs (KG) are far from complete and comprehensive. This problem has motivated efforts in predicting the most plausible missing facts to complete a given KG, i.e., knowledge graph completion (KGC). However, existing KGC methods suffer from two main issues, 1) the false negative issue, i.e., the sampled negative training instances may include potential true facts; and 2) the data sparsity issue, i.e., true facts account for only a tiny part of all possible facts. To this end, we propose positive-unlabeled learning with adversarial data augmentation (PUDA) for KGC. In particular, PUDA tailors positive-unlabeled risk estimator for the KGC task to deal with the false negative issue. Furthermore, to address the data sparsity issue, PUDA achieves a data augmentation strategy by unifying adversarial training and positive-unlabeled learning under the positive-unlabeled minimax game. Extensive experimental results on real-world benchmark datasets demonstrate the effectiveness and compatibility of our proposed method.

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.

MOTIVATION: Protein functions are often described using the Gene Ontology (GO) which is an ontology consisting of over 50 000 classes and a large set of formal axioms. Predicting the functions of proteins is one of the key challenges in computational biology and a variety of machine learning methods have been developed for this purpose. However, these methods usually require a significant amount of training data and cannot make predictions for GO classes that have only few or no experimental annotations. RESULTS: We developed DeepGOZero, a machine learning model which improves predictions for functions with no or only a small number of annotations. To achieve this goal, we rely on a model-theoretic approach for learning ontology embeddings and combine it with neural networks for protein function prediction. DeepGOZero can exploit formal axioms in the GO to make zero-shot predictions, i.e., predict protein functions even if not a single protein in the training phase was associated with that function. Furthermore, the zero-shot prediction method employed by DeepGOZero is generic and can be applied whenever associations with ontology classes need to be predicted. AVAILABILITY AND IMPLEMENTATION: http://github.com/bio-ontology-research-group/deepgozero. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

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.

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.

MOTIVATION: In silico drug-target interaction (DTI) prediction is important for drug discovery and drug repurposing. Approaches to predict DTIs can proceed indirectly, top-down, using phenotypic effects of drugs to identify potential drug targets, or they can be direct, bottom-up and use molecular information to directly predict binding affinities. Both approaches can be combined with information about interaction networks. RESULTS: We developed DTI-Voodoo as a computational method that combines molecular features and ontology-encoded phenotypic effects of drugs with protein-protein interaction networks, and uses a graph convolutional neural network to predict DTIs. We demonstrate that drug effect features can exploit information in the interaction network whereas molecular features do not. DTI-Voodoo is designed to predict candidate drugs for a given protein; we use this formulation to show that common DTI datasets contain intrinsic biases with major effects on performance evaluation and comparison of DTI prediction methods. Using a modified evaluation scheme, we demonstrate that DTI-Voodoo improves significantly over state of the art DTI prediction methods. AVAILABILITY AND IMPLEMENTATION: DTI-Voodoo source code and data necessary to reproduce results are freely available at https://github.com/THinnerichs/DTI-VOODOO. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Motivation: Protein function prediction is one of the major tasks of bioinformatics that can help in wide range of biological problems such as understanding disease mechanisms or finding drug targets.Many methods are available for predicting protein functions from sequence based features, protein-protein interaction networks, protein structure or literature.However, other than sequence, most of the features are difficult to obtain or not available for many proteins thereby limiting their scope.Furthermore, the performance of sequence-based function prediction methods is often lower than methods that incorporate multiple features and predicting protein functions may require a lot of time.Results: We developed a novel method for predicting protein functions from sequence alone which combines deep convolutional neural network (CNN) model with sequence similarity based predictions.Our CNN model scans the sequence for motifs which are predictive for protein functions and combines this with functions of similar proteins (if available).We evaluate the performance of DeepGOPlus using the CAFA3 evaluation measures and achieve an F max of 0.390, 0.557 and 0.614 for BPO, MFO and CCO evaluations, respectively.These results would have made DeepGOPlus one of the three best predictors in CCO and the second best performing method in the BPO and MFO evaluations.We also compare DeepGOPlus with state-of-the-art methods such as DeepText2GO and GOLabeler on another dataset.DeepGOPlus can annotate around 40 protein sequences per second on common hardware, thereby making fast and accurate function predictions available for a wide range of proteins.

MOTIVATION: Infectious diseases caused by novel viruses have become a major public health concern. Rapid identification of virus-host interactions can reveal mechanistic insights into infectious diseases and shed light on potential treatments. Current computational prediction methods for novel viruses are based mainly on protein sequences. However, it is not clear to what extent other important features, such as the symptoms caused by the viruses, could contribute to a predictor. Disease phenotypes (i.e. signs and symptoms) are readily accessible from clinical diagnosis and we hypothesize that they may act as a potential proxy and an additional source of information for the underlying molecular interactions between the pathogens and hosts. RESULTS: We developed DeepViral, a deep learning based method that predicts protein-protein interactions (PPI) between humans and viruses. Motivated by the potential utility of infectious disease phenotypes, we first embedded human proteins and viruses in a shared space using their associated phenotypes and functions, supported by formalized background knowledge from biomedical ontologies. By jointly learning from protein sequences and phenotype features, DeepViral significantly improves over existing sequence-based methods for intra- and inter-species PPI prediction. AVAILABILITY AND IMPLEMENTATION: Code and datasets for reproduction and customization are available at https://github.com/bio-ontology-research-group/DeepViral. Prediction results for 14 virus families are available at https://doi.org/10.5281/zenodo.4429824. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: Protein function prediction is one of the major tasks of bioinformatics that can help in wide range of biological problems such as understanding disease mechanisms or finding drug targets. Many methods are available for predicting protein functions from sequence based features, protein-protein interaction networks, protein structure or literature. However, other than sequence, most of the features are difficult to obtain or not available for many proteins thereby limiting their scope. Furthermore, the performance of sequence-based function prediction methods is often lower than methods that incorporate multiple features and predicting protein functions may require a lot of time. RESULTS: We developed a novel method for predicting protein functions from sequence alone which combines deep convolutional neural network (CNN) model with sequence similarity based predictions. Our CNN model scans the sequence for motifs which are predictive for protein functions and combines this with functions of similar proteins (if available). We evaluate the performance of DeepGOPlus using the CAFA3 evaluation measures and achieve an Fmax of 0.390, 0.557 and 0.614 for BPO, MFO and CCO evaluations, respectively. These results would have made DeepGOPlus one of the three best predictors in CCO and the second best performing method in the BPO and MFO evaluations. We also compare DeepGOPlus with state-of-the-art methods such as DeepText2GO and GOLabeler on another dataset. DeepGOPlus can annotate around 40 protein sequences per second on common hardware, thereby making fast and accurate function predictions available for a wide range of proteins. AVAILABILITY AND IMPLEMENTATION: http://deepgoplus.bio2vec.net/ . SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

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.

Predicting the phenotypes resulting from molecular perturbations is one of the key challenges in genetics. Both forward and reverse genetic screen are employed to identify the molecular mechanisms underlying phenotypes and disease, and these resulted in a large number of genotype-phenotype association being available for humans and model organisms. Combined with recent advances in machine learning, it may now be possible to predict human phenotypes resulting from particular molecular aberrations. We developed DeepPheno, a neural network based hierarchical multi-class multi-label classification method for predicting the phenotypes resulting from loss-of-function in single genes. DeepPheno uses the functional annotations with gene products to predict the phenotypes resulting from a loss-of-function; additionally, we employ a two-step procedure in which we predict these functions first and then predict phenotypes. Prediction of phenotypes is ontology-based and we propose a novel ontology-based classifier suitable for very large hierarchical classification tasks. These methods allow us to predict phenotypes associated with any known protein-coding gene. We evaluate our approach using evaluation metrics established by the CAFA challenge and compare with top performing CAFA2 methods as well as several state of the art phenotype prediction approaches, demonstrating the improvement of DeepPheno over established methods. Furthermore, we show that predictions generated by DeepPheno are applicable to predicting gene-disease associations based on comparing phenotypes, and that a large number of new predictions made by DeepPheno have recently been added as phenotype databases.

An embedding is a function that maps entities from one algebraic structure into another while preserving certain characteristics. Embeddings are being used successfully for mapping relational data or text into vector spaces where they can be used for machine learning, similarity search, or similar tasks. We address the problem of finding vector space embeddings for theories in the Description Logic ??⁺⁺ that are also models of the TBox. To find such embeddings, we define an optimization problem that characterizes the model-theoretic semantics of the operators in ??⁺⁺ within ℝⁿ, thereby solving the problem of finding an interpretation function for an ??⁺⁺ theory given a particular domain Δ. Our approach is mainly relevant to large ??⁺⁺ theories and knowledge bases such as the ontologies and knowledge graphs used in the life sciences. We demonstrate that our method can be used for improved prediction of protein--protein interactions when compared to semantic similarity measures or knowledge graph embeddings.

Entity alignment associates entities in different knowledge graphs if they are semantically same, and has been successfully used in the knowledge graph construction and connection. Most of the recent solutions for entity alignment are based on knowledge graph embedding, which maps knowledge entities in a low-dimension space where entities are connected with the guidance of prior aligned entity pairs. The study in this paper focuses on two important issues that limit the accuracy of current entity alignment solutions: 1) labeled data of priorly aligned entity pairs are difficult and expensive to acquire, whereas abundant of unlabeled data are not used; and 2) knowledge graph embedding is affected by entity's degree difference, which brings challenges to align high frequent and low frequent entities. We propose a semi-supervised entity alignment method (SEA) to leverage both labeled entities and the abundant unlabeled entity information for the alignment. Furthermore, we improve the knowledge graph embedding with awareness of the degree difference by performing the adversarial training. To evaluate our proposed model, we conduct extensive experiments on real-world datasets. The experimental results show that our model consistently outperforms the state-of-the-art methods with significant improvement on alignment accuracy.

MOTIVATION: Ontologies are widely used in biology for data annotation, integration and analysis. In addition to formally structured axioms, ontologies contain meta-data in the form of annotation axioms which provide valuable pieces of information that characterize ontology classes. Annotation axioms commonly used in ontologies include class labels, descriptions or synonyms. Despite being a rich source of semantic information, the ontology meta-data are generally unexploited by ontology-based analysis methods such as semantic similarity measures. RESULTS: We propose a novel method, OPA2Vec, to generate vector representations of biological entities in ontologies by combining formal ontology axioms and annotation axioms from the ontology meta-data. We apply a Word2Vec model that has been pre-trained on either a corpus or abstracts or full-text articles to produce feature vectors from our collected data. We validate our method in two different ways: first, we use the obtained vector representations of proteins in a similarity measure to predict protein-protein interaction on two different datasets. Second, we evaluate our method on predicting gene-disease associations based on phenotype similarity by generating vector representations of genes and diseases using a phenotype ontology, and applying the obtained vectors to predict gene-disease associations using mouse model phenotypes. We demonstrate that OPA2Vec significantly outperforms existing methods for predicting gene-disease associations. Using evidence from mouse models, we apply OPA2Vec to identify candidate genes for several thousand rare and orphan diseases. OPA2Vec can be used to produce vector representations of any biomedical entity given any type of biomedical ontology. AVAILABILITY AND IMPLEMENTATION: https://github.com/bio-ontology-research-group/opa2vec. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Prioritization of variants in personal genomic data is a major challenge. Recently, computational methods that rely on comparing phenotype similarity have shown to be useful to identify causative variants. In these methods, pathogenicity prediction is combined with a semantic similarity measure to prioritize not only variants that are likely to be dysfunctional but those that are likely involved in the pathogenesis of a patient's phenotype. RESULTS: We have developed DeepPVP, a variant prioritization method that combined automated inference with deep neural networks to identify the likely causative variants in whole exome or whole genome sequence data. We demonstrate that DeepPVP performs significantly better than existing methods, including phenotype-based methods that use similar features. DeepPVP is freely available at https://github.com/bio-ontology-research-group/phenomenet-vp . CONCLUSIONS: DeepPVP further improves on existing variant prioritization methods both in terms of speed as well as accuracy.

The past decade has seen an explosion in the collection of genotype data in domains as diverse as medicine, ecology, livestock and plant breeding. Along with this comes the challenge of dealing with the related phenotype data, which is not only large but also highly multidimensional. Computational analysis of phenotypes has therefore become critical for our ability to understand the biological meaning of genomic data in the biological sciences. At the heart of computational phenotype analysis are the phenotype ontologies. A large number of these ontologies have been developed across many domains, and we are now at a point where the knowledge captured in the structure of these ontologies can be used for the integration and analysis of large interrelated data sets. The Phenotype And Trait Ontology framework provides a method for formal definitions of phenotypes and associated data sets and has proved to be key to our ability to develop methods for the integration and analysis of phenotype data. Here, we describe the development and products of the ontological approach to phenotype capture, the formal content of phenotype ontologies and how their content can be used computationally.

Systems biology models are rapidly increasing in complexity, size and numbers. When building large models, researchers rely on software tools for the retrieval, comparison, combination and merging of models, as well as for version control. These tools need to be able to quantify the differences and similarities between computational models. However, depending on the specific application, the notion of 'similarity' may greatly vary. A general notion of model similarity, applicable to various types of models, is still missing. Here we survey existing methods for the comparison of models, introduce quantitative measures for model similarity, and discuss potential applications of combined similarity measures. To frame model comparison as a general problem, we describe a theoretical approach to defining and computing similarities based on a combination of different model aspects. The six aspects that we define as potentially relevant for similarity are underlying encoding, references to biological entities, quantitative behaviour, qualitative behaviour, mathematical equations and parameters and network structure. We argue that future similarity measures will benefit from combining these model aspects in flexible, problem-specific ways to mimic users' intuition about model similarity, and to support complex model searches in databases.

Motivation: In the past years, several methods have been developed to incorporate information about phenotypes into computational disease gene prioritization methods. These methods commonly compute the similarity between a disease's (or patient's) phenotypes and a database of gene-to-phenotype associations to find the phenotypically most similar match. A key limitation of these methods is their reliance on knowledge about phenotypes associated with particular genes which is highly incomplete in humans as well as in many model organisms such as the mouse. Results: We developed SmuDGE, a method that uses feature learning to generate vector-based representations of phenotypes associated with an entity. SmuDGE can be used as a trainable semantic similarity measure to compare two sets of phenotypes (such as between a disease and gene, or a disease and patient). More importantly, SmuDGE can generate phenotype representations for entities that are only indirectly associated with phenotypes through an interaction network; for this purpose, SmuDGE exploits background knowledge in interaction networks comprised of multiple types of interactions. We demonstrate that SmuDGE can match or outperform semantic similarity in phenotype-based disease gene prioritization, and furthermore significantly extends the coverage of phenotype-based methods to all genes in a connected interaction network. Availability and implementation: https://github.com/bio-ontology-research-group/SmuDGE.

Motivation: A large number of protein sequences are becoming available through the application of novel high-throughput sequencing technologies. Experimental functional characterization of these proteins is time-consuming and expensive, and is often only done rigorously for few selected model organisms. Computational function prediction approaches have been suggested to fill this gap. The functions of proteins are classified using the Gene Ontology (GO), which contains over 40 000 classes. Additionally, proteins have multiple functions, making function prediction a large-scale, multi-class, multi-label problem. Results: We have developed a novel method to predict protein function from sequence. We use deep learning to learn features from protein sequences as well as a cross-species protein-protein interaction network. Our approach specifically outputs information in the structure of the GO and utilizes the dependencies between GO classes as background information to construct a deep learning model. We evaluate our method using the standards established by the Computational Assessment of Function Annotation (CAFA) and demonstrate a significant improvement over baseline methods such as BLAST, in particular for predicting cellular locations. Availability and implementation: Web server: http://deepgo.bio2vec.net, Source code: https://github.com/bio-ontology-research-group/deepgo. Contact: [email protected]. Supplementary information: Supplementary data are available at Bioinformatics online.

Motivation: Function annotations of gene products, and phenotype annotations of genotypes, provide valuable information about molecular mechanisms that can be utilized by computational methods to identify functional and phenotypic relatedness, improve our understanding of disease and pathobiology, and lead to discovery of drug targets. Identifying functions and phenotypes commonly requires experiments which are time-consuming and expensive to carry out; creating the annotations additionally requires a curator to make an assertion based on reported evidence. Support to validate the mutual consistency of functional and phenotype annotations as well as a computational method to predict phenotypes from function annotations, would greatly improve the utility of function annotations. Results: We developed a novel ontology-based method to validate the mutual consistency of function and phenotype annotations. We apply our method to mouse and human annotations, and identify several inconsistencies that can be resolved to improve overall annotation quality. We also apply our method to the rule-based prediction of regulatory phenotypes from functions and demonstrate that we can predict these phenotypes with Fmax of up to 0.647. Availability and implementation: https://github.com/bio-ontology-research-group/phenogocon.

Motivation: Biological knowledge is widely represented in the form of ontology-based annotations: ontologies describe the phenomena assumed to exist within a domain, and the annotations associate a (kind of) biological entity with a set of phenomena within the domain. The structure and information contained in ontologies and their annotations make them valuable for developing machine learning, data analysis and knowledge extraction algorithms; notably, semantic similarity is widely used to identify relations between biological entities, and ontology-based annotations are frequently used as features in machine learning applications. Results: We propose the Onto2Vec method, an approach to learn feature vectors for biological entities based on their annotations to biomedical ontologies. Our method can be applied to a wide range of bioinformatics research problems such as similarity-based prediction of interactions between proteins, classification of interaction types using supervised learning, or clustering. To evaluate Onto2Vec, we use the gene ontology (GO) and jointly produce dense vector representations of proteins, the GO classes to which they are annotated, and the axioms in GO that constrain these classes. First, we demonstrate that Onto2Vec-generated feature vectors can significantly improve prediction of protein-protein interactions in human and yeast. We then illustrate how Onto2Vec representations provide the means for constructing data-driven, trainable semantic similarity measures that can be used to identify particular relations between proteins. Finally, we use an unsupervised clustering approach to identify protein families based on their Enzyme Commission numbers. Our results demonstrate that Onto2Vec can generate high quality feature vectors from biological entities and ontologies. Onto2Vec has the potential to significantly outperform the state-of-the-art in several predictive applications in which ontologies are involved. Availability and implementation: https://github.com/bio-ontology-research-group/onto2vec. Supplementary information: Supplementary data are available at Bioinformatics online.

BACKGROUND: Ontologies are representations of a conceptualization of a domain. Traditionally, ontologies in biology were represented as directed acyclic graphs (DAG) which represent the backbone taxonomy and additional relations between classes. These graphs are widely exploited for data analysis in the form of ontology enrichment or computation of semantic similarity. More recently, ontologies are developed in a formal language such as the Web Ontology Language (OWL) and consist of a set of axioms through which classes are defined or constrained. While the taxonomy of an ontology can be inferred directly from the axioms of an ontology as one of the standard OWL reasoning tasks, creating general graph structures from OWL ontologies that exploit the ontologies' semantic content remains a challenge. RESULTS: We developed a method to transform ontologies into graphs using an automated reasoner while taking into account all relations between classes. Searching for (existential) patterns in the deductive closure of ontologies, we can identify relations between classes that are implied but not asserted and generate graph structures that encode for a large part of the ontologies' semantic content. We demonstrate the advantages of our method by applying it to inference of protein-protein interactions through semantic similarity over the Gene Ontology and demonstrate that performance is increased when graph structures are inferred using deductive inference according to our method. Our software and experiment results are available at http://github.com/bio-ontology-research-group/Onto2Graph . CONCLUSIONS: Onto2Graph is a method to generate graph structures from OWL ontologies using automated reasoning. The resulting graphs can be used for improved ontology visualization and ontology-based data analysis.

Briefings in Bioinformatics (2016). doi: 10.1093/bib/bbw090 The authors of the above article wish to correct the affiliation for Tim Kacprowski and the funding section for clarity. The corrections have been made online and in print. The authors apologize for this error.

MOTIVATION: 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. AVAILABILITY AND IMPLEMENTATION: https://github.com/bio-ontology-research-group/walking-rdf-and-owl. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

BACKGROUND: Cell lines and cell types are extensively studied in biomedical research yielding to a significant amount of publications each year. Identifying cell lines and cell types precisely in publications is crucial for science reproducibility and knowledge integration. There are efforts for standardisation of the cell nomenclature based on ontology development to support FAIR principles of the cell knowledge. However, it is important to analyse the usage of cell nomenclature in publications at a large scale for understanding the level of uptake of cell nomenclature in literature by scientists. In this study, we analyse the usage of cell nomenclature, both in Vivo, and in Vitro in biomedical literature by using text mining methods and present our results. RESULTS: We identified 59% of the cell type classes in the Cell Ontology and 13% of the cell line classes in the Cell Line Ontology in the literature. Our analysis showed that cell line nomenclature is much more ambiguous compared to the cell type nomenclature. However, trends indicate that standardised nomenclature for cell lines and cell types are being increasingly used in publications by the scientists. CONCLUSIONS: Our findings provide an insight to understand how experimental cells are described in publications and may allow for an improved standardisation of cell type and cell line nomenclature as well as can be utilised to develop efficient text mining applications on cell types and cell lines. All data generated in this study is available at https://github.com/shenay/CellNomenclatureStudy.

Discriminating the causative disease variant(s) for individuals with inherited or de novo mutations presents one of the main challenges faced by the clinical genetics community today. Computational approaches for variant prioritization include machine learning methods utilizing a large number of features, including molecular information, interaction networks, or phenotypes. Here, we demonstrate the PhenomeNET Variant Predictor (PVP) system that exploits semantic technologies and automated reasoning over genotype-phenotype relations to filter and prioritize variants in whole exome and whole genome sequencing datasets. We demonstrate the performance of PVP in identifying causative variants on a large number of synthetic whole exome and whole genome sequences, covering a wide range of diseases and syndromes. In a retrospective study, we further illustrate the application of PVP for the interpretation of whole exome sequencing data in patients suffering from congenital hypothyroidism. We find that PVP accurately identifies causative variants in whole exome and whole genome sequencing datasets and provides a powerful resource for the discovery of causal variants.

A large number of biomedical resources have been developed to represent the functions of biological entities, and these resources are widely used for data integration and analysis. Expressing functions in biomedical ontologies currently uses formal representation patterns that renders basic reasoning tasks to fall in complexity classes beyond polynomial time, thereby limiting the potential of using knowledge-based methods for data integration, querying or quality control. Here, we propose an alternative representation pattern for expressing knowledge about biological functions, together with a biological and ontological justification, which can be expressed using the description logic EL++ and implemented using the OWL 2 EL profile. To demonstrate the utility of our account of biological functions, we apply it to all proteins contained in the SwissProt database and evaluate its utility with respect to answering complex queries as well with respect to the classification and query times.

Ontologies are widely used in biological and biomedical research. Their success lies in their combination of four main features present in almost all ontologies: provision of standard identifiers for classes and relations that represent the phenomena within a domain; provision of a vocabulary for a domain; provision of metadata that describes the intended meaning of the classes and relations in ontologies; and the provision of machine-readable axioms and definitions that enable computational access to some aspects of the meaning of classes and relations. While each of these features enables applications that facilitate data integration, data access and analysis, a great potential lies in the possibility of combining these four features to support integrative analysis and interpretation of multimodal data. Here, we provide a functional perspective on ontologies in biology and biomedicine, focusing on what ontologies can do and describing how they can be used in support of integrative research. We also outline perspectives for using ontologies in data-driven science, in particular their application in structured data mining and machine learning applications.

BACKGROUND: Many ontologies have been developed in biology and these ontologies increasingly contain large volumes of formalized knowledge commonly expressed in the Web Ontology Language (OWL). Computational access to the knowledge contained within these ontologies relies on the use of automated reasoning. RESULTS: We have developed the Aber-OWL infrastructure that provides reasoning services for bio-ontologies. Aber-OWL consists of an ontology repository, a set of web services and web interfaces that enable ontology-based semantic access to biological data and literature. Aber-OWL is freely available at http://aber-owl.net . CONCLUSIONS: Aber-OWL provides a framework for automatically accessing information that is annotated with ontologies or contains terms used to label classes in ontologies. When using Aber-OWL, access to ontologies and data annotated with them is not merely based on class names or identifiers but rather on the knowledge the ontologies contain and the inferences that can be drawn from it.

MOTIVATION: Methods for computational drug target identification use information from diverse information sources to predict or prioritize drug targets for known drugs. One set of resources that has been relatively neglected for drug repurposing is animal model phenotype. RESULTS: We investigate the use of mouse model phenotypes for drug target identification. To achieve this goal, we first integrate mouse model phenotypes and drug effects, and then systematically compare the phenotypic similarity between mouse models and drug effect profiles. We find a high similarity between phenotypes resulting from loss-of-function mutations and drug effects resulting from the inhibition of a protein through a drug action, and demonstrate how this approach can be used to suggest candidate drug targets. AVAILABILITY AND IMPLEMENTATION: Analysis code and supplementary data files are available on the project Web site at https://drugeffects.googlecode.com.

Ontologies are now pervasive in biomedicine, where they serve as a means to standardize terminology, to enable access to domain knowledge, to verify data consistency and to facilitate integrative analyses over heterogeneous biomedical data. For this purpose, research on biomedical ontologies applies theories and methods from diverse disciplines such as information management, knowledge representation, cognitive science, linguistics and philosophy. Depending on the desired applications in which ontologies are being applied, the evaluation of research in biomedical ontologies must follow different strategies. Here, we provide a classification of research problems in which ontologies are being applied, focusing on the use of ontologies in basic and translational research, and we demonstrate how research results in biomedical ontologies can be evaluated. The evaluation strategies depend on the desired application and measure the success of using an ontology for a particular biomedical problem. For many applications, the success can be quantified, thereby facilitating the objective evaluation and comparison of research in biomedical ontology. The objective, quantifiable comparison of research results based on scientific applications opens up the possibility for systematically improving the utility of ontologies in biomedical research.

MOTIVATION: Many complex diseases are the result of abnormal pathway functions instead of single abnormalities. Disease diagnosis and intervention strategies must target these pathways while minimizing the interference with normal physiological processes. Large-scale identification of disease pathways and chemicals that may be used to perturb them requires the integration of information about drugs, genes, diseases and pathways. This information is currently distributed over several pharmacogenomics databases. An integrated analysis of the information in these databases can reveal disease pathways and facilitate novel biomedical analyses. RESULTS: We demonstrate how to integrate pharmacogenomics databases through integration of the biomedical ontologies that are used as meta-data in these databases. The additional background knowledge in these ontologies can then be used to enable novel analyses. We identify disease pathways using a novel multi-ontology enrichment analysis over the Human Disease Ontology, and we identify significant associations between chemicals and pathways using an enrichment analysis over a chemical ontology. The drug-pathway and disease-pathway associations are a valuable resource for research in disease and drug mechanisms and can be used to improve computational drug repurposing. AVAILABILITY: http://pharmgkb-owl.googlecode.com

MOTIVATION: The systematic observation of phenotypes has become a crucial tool of functional genomics, and several large international projects are currently underway to identify and characterize the phenotypes that are associated with genotypes in several species. To integrate phenotype descriptions within and across species, phenotype ontologies have been developed. Applying ontologies to unify phenotype descriptions in the domain of physiology has been a particular challenge due to the high complexity of the underlying domain. RESULTS: In this study, we present the outline of a theory and its implementation for an ontology of physiology-related phenotypes. We provide a formal description of process attributes and relate them to the attributes of their temporal parts and participants. We apply our theory to create the Cellular Phenotype Ontology (CPO). The CPO is an ontology of morphological and physiological phenotypic characteristics of cells, cell components and cellular processes. Its prime application is to provide terms and uniform definition patterns for the annotation of cellular phenotypes. The CPO can be used for the annotation of observed abnormalities in domains, such as systems microscopy, in which cellular abnormalities are observed and for which no phenotype ontology has been created. AVAILABILITY AND IMPLEMENTATION: The CPO and the source code we generated to create the CPO are freely available on http://cell-phenotype.googlecode.com.

SUMMARY: The article presents an infrastructure for supporting the semantic interoperability of biomedical resources based on the management (storing and inference-based querying) of their ontology-based annotations. This infrastructure consists of: (i) a repository to store and query ontology-based annotations; (ii) a knowledge base server with an inference engine to support the storage of and reasoning over ontologies used in the annotation of resources; (iii) a set of applications and services allowing interaction with the integrated repository and knowledge base. The infrastructure is being prototyped and developed and evaluated by the RICORDO project in support of the knowledge management of biomedical resources, including physiology and pharmacology models and associated clinical data. AVAILABILITY AND IMPLEMENTATION: The RICORDO toolkit and its source code are freely available from http://ricordo.eu/relevant-resources. CONTACT: [email protected].

Abstract Motivation: Primary immunodeficiency diseases (PIDs) are Mendelian conditions of high phenotypic complexity and low incidence. They usually manifest in toddlers and infants, although they can also occur much later in life. Information about PIDs is often widely scattered throughout the clinical as well as the research literature and hard to find for both generalists as well as experienced clinicians. Semantic Web technologies coupled to clinical information systems can go some way toward addressing this problem. Ontologies are a central component of such a system, containing and centralizing knowledge about primary immunodeficiencies in both a human- and computer-comprehensible form. The development of an ontology of PIDs is therefore a central step toward developing informatics tools, which can support the clinician in the diagnosis and treatment of these diseases. Results: We present PIDO, the primary immunodeficiency disease ontology. PIDO characterizes PIDs in terms of the phenotypes commonly observed by clinicians during a diagnosis process. Phenotype terms in PIDO are formally defined using complex definitions based on qualities, functions, processes and structures. We provide mappings to biomedical reference ontologies to ensure interoperability with ontologies in other domains. Based on PIDO, we developed the PIDFinder, an ontology-driven software prototype that can facilitate clinical decision support. PIDO connects immunological knowledge across resources within a common framework and thereby enables translational research and the development of medical applications for the domain of immunology and primary immunodeficiency diseases. Availability: The Primary Immunodeficiency Disease Ontology is available under a Creative Commons Attribution 3.0 (CC-BY 3.0) licence at http://code.google.com/p/pido/. The most recent public release of the ontology can always be found at http://purl.org/scimantica/pido/owl/pid.owl. An instance of the PIDFinder software can be found at http://pidfinder.appspot.com Contact: [email protected]

MOTIVATION: Ontologies are essential in biomedical research due to their ability to semantically integrate content from different scientific databases and resources. Their application improves capabilities for querying and mining biological knowledge. An increasing number of ontologies is being developed for this purpose, and considerable effort is invested into formally defining them in order to represent their semantics explicitly. However, current biomedical ontologies do not facilitate data integration and interoperability yet, since reasoning over these ontologies is very complex and cannot be performed efficiently or is even impossible. We propose the use of less expressive subsets of ontology representation languages to enable efficient reasoning and achieve the goal of genuine interoperability between ontologies. RESULTS: We present and evaluate EL Vira, a framework that transforms OWL ontologies into the OWL EL subset, thereby enabling the use of tractable reasoning. We illustrate which OWL constructs and inferences are kept and lost following the conversion and demonstrate the performance gain of reasoning indicated by the significant reduction of processing time. We applied EL Vira to the open biomedical ontologies and provide a repository of ontologies resulting from this conversion. EL Vira creates a common layer of ontological interoperability that, for the first time, enables the creation of software solutions that can employ biomedical ontologies to perform inferences and answer complex queries to support scientific analyses. AVAILABILITY AND IMPLEMENTATION: The EL Vira software is available from http://el-vira.googlecode.com and converted OBO ontologies and their mappings are available from http://bioonto.gen.cam.ac.uk/el-ont.

Science aims to develop an accurate understanding of reality through a variety of rigorously empirical and formal methods. Ontologies are used to formalize the meaning of terms within a domain of discourse. The Basic Formal Ontology (BFO) is an ontology of particular importance in the biomedical domains, where it provides the top-level for numerous ontologies, including those admitted as part of the OBO Foundry collection. The BFO requires that all classes in an ontology are actually instantiated in reality. Despite the fact that it is hard to show whether entities of some kind exist or do not exist in reality (especially for unobservable entities like elementary particles), this criterion fails to satisfy the need of scientists to communicate their findings and theories unambiguously. We discuss the problems that arise due to the BFO's realism criterion and suggest viable alternatives.

MOTIVATION: Phenotypic information is important for the analysis of the molecular mechanisms underlying disease. A formal ontological representation of phenotypic information can help to identify, interpret and infer phenotypic traits based on experimental findings. The methods that are currently used to represent data and information about phenotypes fail to make the semantics of the phenotypic trait explicit and do not interoperate with ontologies of anatomy and other domains. Therefore, valuable resources for the analysis of phenotype studies remain unconnected and inaccessible to automated analysis and reasoning. RESULTS: We provide a framework to formalize phenotypic descriptions and make their semantics explicit. Based on this formalization, we provide the means to integrate phenotypic descriptions with ontologies of other domains, in particular anatomy and physiology. We demonstrate how our framework leads to the capability to represent disease phenotypes, perform powerful queries that were not possible before and infer additional knowledge. AVAILABILITY: http://bioonto.de/pmwiki.php/Main/PheneOntology.

BACKGROUND: most biomedical ontologies are represented in the OBO Flatfile Format, which is an easy-to-use graph-based ontology language. The semantics of the OBO Flatfile Format 1.2 enforces a strict predetermined interpretation of relationship statements between classes. It does not allow flexible specifications that provide better approximations of the intuitive understanding of the considered relations. If relations cannot be accurately expressed then ontologies built upon them may contain false assertions and hence lead to false inferences. Ontologies in the OBO Foundry must formalize the semantics of relations according to the OBO Relationship Ontology (RO). Therefore, being able to accurately express the intended meaning of relations is of crucial importance. Since the Web Ontology Language (OWL) is an expressive language with a formal semantics, it is suitable to de ne the meaning of relations accurately. RESULTS: we developed a method to provide definition patterns for relations between classes using OWL and describe a novel implementation of the RO based on this method. We implemented our extension in software that converts ontologies in the OBO Flatfile Format to OWL, and also provide a prototype to extract relational patterns from OWL ontologies using automated reasoning. The conversion software is freely available at http://bioonto.de/obo2owl, and can be accessed via a web interface. CONCLUSIONS: explicitly defining relations permits their use in reasoning software and leads to a more flexible and powerful way of representing biomedical ontologies. Using the extended langua0067e and semantics avoids several mistakes commonly made in formalizing biomedical ontologies, and can be used to automatically detect inconsistencies. The use of our method enables the use of graph-based ontologies in OWL, and makes complex OWL ontologies accessible in a graph-based form. Thereby, our method provides the means to gradually move the representation of biomedical ontologies into formal knowledge representation languages that incorporates an explicit semantics. Our method facilitates the use of OWL-based software in the back-end while ontology curators may continue to develop ontologies with an OBO-style front-end.

The development and verification of software models that are applicable across multiple domains remains a difficult problem. We propose a novel approach to model-driven software development based on ontologies and Semantic Web technology. Our approach uses three ontologies to define software models: a task ontology, a domain ontology and a top-level ontology. The task ontology serves as the conceptual model for the software, the domain ontology provides domain-specific knowledge and the top-level ontology integrates the task and domain ontologies. Our method allows the verification of these models both for consistency and ontological adequacy. This verification can be performed both at development and runtime. Domain ontologies are replaceable modules, which enables the comparison and application of the models built using our method across multiple domains. We demonstrate the viability of our approach through the design and implementation of a semantic wiki and a social tagging system, and compare it with model-driven software development to illustrate its benefits.

MOTIVATION: Ontology development and the annotation of biological data using ontologies are time-consuming exercises that currently require input from expert curators. Open, collaborative platforms for biological data annotation enable the wider scientific community to become involved in developing and maintaining such resources. However, this openness raises concerns regarding the quality and correctness of the information added to these knowledge bases. The combination of a collaborative web-based platform with logic-based approaches and Semantic Web technology can be used to address some of these challenges and concerns. RESULTS: We have developed the BOWiki, a web-based system that includes a biological core ontology. The core ontology provides background knowledge about biological types and relations. Against this background, an automated reasoner assesses the consistency of new information added to the knowledge base. The system provides a platform for research communities to integrate information and annotate data collaboratively. AVAILABILITY: The BOWiki and supplementary material is available at http://www.bowiki.net/. The source code is available under the GNU GPL from http://onto.eva.mpg.de/trac/BoWiki.

BACKGROUND: Biological sequences play a major role in molecular and computational biology. They are studied as information-bearing entities that make up DNA, RNA or proteins. The Sequence Ontology, which is part of the OBO Foundry, contains descriptions and definitions of sequences and their properties. Yet the most basic question about sequences remains unanswered: what kind of entity is a biological sequence? An answer to this question benefits formal ontologies that use the notion of biological sequences and analyses in computational biology alike. RESULTS: We provide both an ontological analysis of biological sequences and a formal representation that can be used in knowledge-based applications and other ontologies. We distinguish three distinct kinds of entities that can be referred to as "biological sequence": chains of molecules, syntactic representations such as those in biological databases, and the abstract information-bearing entities. For use in knowledge-based applications and inclusion in biomedical ontologies, we implemented the developed axiom system for use in automated theorem proving. CONCLUSION: Axioms are necessary to achieve the main goal of ontologies: to formally specify the meaning of terms used within a domain. The axiom system for the ontology of biological sequences is the first elaborate axiom system for an OBO Foundry ontology and can serve as starting point for the development of more formal ontologies and ultimately of knowledge-based applications.

BACKGROUND: Current efforts within the biomedical ontology community focus on achieving interoperability between various biomedical ontologies that cover a range of diverse domains. Achieving this interoperability will contribute to the creation of a rich knowledge base that can be used for querying, as well as generating and testing novel hypotheses. The OBO Foundry principles, as applied to a number of biomedical ontologies, are designed to facilitate this interoperability. However, semantic extensions are required to meet the OBO Foundry interoperability goals. Inconsistencies may arise when ontologies of properties - mostly phenotype ontologies - are combined with ontologies taking a canonical view of a domain - such as many anatomical ontologies. Currently, there is no support for a correct and consistent integration of such ontologies. RESULTS: We have developed a methodology for accurately representing canonical domain ontologies within the OBO Foundry. This is achieved by adding an extension to the semantics for relationships in the biomedical ontologies that allows for treating canonical information as default. Conclusions drawn from default knowledge may be revoked when additional information becomes available. We show how this extension can be used to achieve interoperability between ontologies, and further allows for the inclusion of more knowledge within them. We apply the formalism to ontologies of mouse anatomy and mammalian phenotypes in order to demonstrate the approach. CONCLUSION: Biomedical ontologies require a new class of relations that can be used in conjunction with default knowledge, thereby extending those currently in use. The inclusion of default knowledge is necessary in order to ensure interoperability between ontologies.