We propose the molecular omics network (MOOMIN) a multimodal graph neural network used by AstraZeneca oncologists to predict the synergy of drug combinations for cancer treatment. Our model learns drug representations at multiple scales based on a drug-protein interaction network and metadata. Structural properties of compounds and proteins are encoded to create vertex features for a message-passing scheme that operates on the bipartite interaction graph. Propagated messages form multi-resolution drug representations which we utilized to create drug pair descriptors. By conditioning the drug combination representations on the cancer cell type we define a synergy scoring function that can inductively score unseen pairs of drugs. Experimental results on the synergy scoring task demonstrate that MOOMIN outperforms state-of-the-art graph fingerprinting, proximity preserving node embedding, and existing deep learning approaches. Further results establish that the predictive performance of our model is robust to hyperparameter changes. We demonstrate that the model makes high-quality predictions over a wide range of cancer cell line tissues, out-of-sample predictions can be validated with external synergy databases, and that the proposed model is data efficient at learning.

In recent years, numerous machine learning models which attempt to solve polypharmacy side effect identification, drug-drug interaction prediction, and combination therapy design tasks have been proposed. Here, we present a unified theoretical view of relational machine learning models which can address these tasks. We provide fundamental definitions, compare existing model architectures and discuss performance metrics, datasets, and evaluation protocols. In addition, we emphasize possible high-impact applications and important future research directions in this domain.

In this paper, we introduce ChemicalX, a PyTorch-based deep learning library designed for providing a range of state of the art models to solve the drug pair scoring task. The primary objective of the library is to make deep drug pair scoring models accessible to machine learning researchers and practitioners in a streamlined framework. The design of ChemicalX reuses existing high level model training utilities, geometric deep learning, and deep chemistry layers from the PyTorch ecosystem. Our system provides neural network layers, custom pair scoring architectures, data loaders, and batch iterators for end users. We showcase these features with example code snippets and case studies to highlight the characteristics of ChemicalX. A range of experiments on real world drug-drug interaction, polypharmacy side effect, and combination synergy prediction tasks demonstrate that the models available in ChemicalX are effective at solving the pair scoring task. Finally, we show that ChemicalX could be used to train and score machine learning models on large drug pair datasets with hundreds of thousands of compounds on commodity hardware.

While neural networks models have shown impressive performance in many NLP tasks, lack of interpretability is often seen as a disadvantage. Individual relevance scores assigned by post-hoc explanation methods are not sufficient to show deeper systematic preferences and potential biases of the model that apply consistently across examples. In this paper we apply rule mining using knowledge graphs in combination with neural network explanation methods to uncover such systematic preferences of trained neural models and capture them in the form of conjunctive rules. We test our approach in the context of text classification tasks and show that such rules are able to explain a substantial part of the model behaviour as well as indicate potential causes of misclassifications when the model is applied outside of the initial training context.

One of the key interests for learning analytics is how it can be used to improve retention. This paper focuses on work conducted at the Open University (OU) into predicting students who are at risk of failing their module. The Open University is one of the worlds largest distance learning institutions. Since tutors do not interact face to face with students, it can be difficult for tutors to identify and respond to students who are struggling in time to try to resolve the difficulty. Predictive models have been developed and tested using historic Virtual Learning Environment (VLE) activity data combined with other data sources, for three OU modules. This has revealed that it is possible to predict student failure by looking for changes in user's activity in the VLE, when compared against their own previous behaviour, or that of students who can be categorised as having similar learning behaviour. More focused analysis of these modules applying the GUHA (General Unary Hypothesis Automaton) method of data analysis has also yielded some early promising results for creating accurate hypothesis about students who fail.

The W3C Semantic Sensor Network Incubator group (the SSN-XG) produced an OWL 2 ontology to describe sensors and observations — the SSN ontology, available at http://purl.oclc.org/NET/ssnx/ssn. The SSN ontology can describe sensors in terms of capabilities, measurement processes, observations and deployments. This article describes the SSN ontology. It further gives an example and describes the use of the ontology in recent research projects.

By specifying that published datasets must link to other existing datasets, the 4th linked data principle ensures a Web of data and not just a set of unconnected data islands. The authors propose in this paper the term data linking to name the problem of finding equivalent resources on the Web of linked data. In order to perform data linking, many techniques were developed, finding their roots in statistics, database, natural language processing and graph theory. The authors begin this paper by providing background information and terminological clarifications related to data linking. Then a comprehensive survey over the various techniques available for data linking is provided. These techniques are classified along the three criteria of granularity, type of evidence, and source of the evidence. Finally, the authors survey eleven recent tools performing data linking and we classify them according to the surveyed techniques.

We propose a knowledge fusion architecture KnoFuss based on the application of problem-solving methods technology, which allows methods for subtasks of the fusion process to be combined and the best methods to be selected, depending on the domain and task at hand.