Search

Summary Knowledge graphs (KGs) are powerful tools for representing and organizing complex biomedical data. They empower researchers, physicians, and scientists by facilitating rapid access to biomedical information, enabling the discernment of patterns or insights, and fostering the formulation of decisions and the generation of novel knowledge. To automate these activities, several KG embedding algorithms have been proposed to learn from and complete KGs. However, the efficacy of these embedding algorithms appears limited when applied to biomedical KGs, prompting questions about whether they can be useful in this field. To that end, we explore several widely used KG embedding models and evaluate their performance and applications using a recent biomedical KG, BioKG. We also demonstrate that by using recent best practices for training KG embeddings, it is possible to improve performance over BioKG. Additionally, we address interpretability concerns that naturally arise with such machine learning methods. In particular, we examine rule-based methods that aim to address these concerns by making interpretable predictions using learned rules, achieving comparable performance. Finally, we discuss a realistic use case where a pretrained BioKG embedding is further trained for a specific task, in this case, four polypharmacy scenarios where the goal is to predict missing links or entities in another downstream KGs in four polypharmacy scenarios. We conclude that in the right scenarios, biomedical KG embeddings can be effective and useful. Availability and implementation Our code and data is available at https://github.com/aryopg/biokge. [ABSTRACT FROM AUTHOR]

This article summarizes principles and ideas from the emerging area of applying conditional computation methods to the design of neural networks. In particular, we focus on neural networks that can dynamically activate or de-activate parts of their computational graph conditionally on their input. Examples include the dynamic selection of, e.g., input tokens, layers (or sets of layers), and sub-modules inside each layer (e.g., channels in a convolutional filter). We first provide a general formalism to describe these techniques in an uniform way. Then, we introduce three notable implementations of these principles: mixture-of-experts (MoEs) networks, token selection mechanisms, and early-exit neural networks. The paper aims to provide a tutorial-like introduction to this growing field. To this end, we analyze the benefits of these modular designs in terms of efficiency, explainability, and transfer learning, with a focus on emerging applicative areas ranging from automated scientific discovery to semantic communication. [ABSTRACT FROM AUTHOR]

PURPOSE: Stratifying patients with cancer according to risk of relapse can personalize their care. In this work, we provide an answer to the following research question: How to use machine learning to estimate probability of relapse in patients with early-stage non–small-cell lung cancer (NSCLC)? MATERIALS AND METHODS: For predicting relapse in 1,387 patients with early-stage (I-II) NSCLC from the Spanish Lung Cancer Group data (average age 65.7 years, female 24.8%, male 75.2%), we train tabular and graph machine learning models. We generate automatic explanations for the predictions of such models. For models trained on tabular data, we adopt SHapley Additive exPlanations local explanations to gauge how each patient feature contributes to the predicted outcome. We explain graph machine learning predictions with an example-based method that highlights influential past patients. RESULTS: Machine learning models trained on tabular data exhibit a 76% accuracy for the random forest model at predicting relapse evaluated with a 10-fold cross-validation (the model was trained 10 times with different independent sets of patients in test, train, and validation sets, and the reported metrics are averaged over these 10 test sets). Graph machine learning reaches 68% accuracy over a held-out test set of 200 patients, calibrated on a held-out set of 100 patients. CONCLUSION: Our results show that machine learning models trained on tabular and graph data can enable objective, personalized, and reproducible prediction of relapse and, therefore, disease outcome in patients with early-stage NSCLC. With further prospective and multisite validation, and additional radiological and molecular data, this prognostic model could potentially serve as a predictive decision support tool for deciding the use of adjuvant treatments in early-stage lung cancer. [ABSTRACT FROM AUTHOR]

We focus on the problem of predicting missing assertions in Web ontologies. We start from the assumption that individual resources that are similar in some aspects are more likely to be linked by specific relations: this phenomenon is also referred to as homophily and emerges in a variety of relational domains. In this article, we propose a method for (1) identifying which relations in the ontology are more likely to link similar individuals and (2) efficiently propagating knowledge across chains of similar individuals. By enforcing sparsity in the model parameters, the proposed method is able to select only the most relevant relations for a given prediction task. Our experimental evaluation demonstrates the effectiveness of the proposed method in comparison to state-of-the-art methods from the literature. [ABSTRACT FROM AUTHOR]

We focus on the problem of link prediction in Knowledge Graphs, with the goal of discovering new facts. To this purpose, Energy-Based Models for Knowledge Graphs that embed entities and relations in continuous vector spaces have been largely used. The main limitation in their applicability lies in the parameter learning phase, which may require a large amount of time for converging to optimal solutions. In this article, we first propose an unified view on different Energy-Based Embedding Models. Hence, for improving the model training phase, we propose the adoption of adaptive learning rates. We show that, by adopting adaptive learning rates during training, we can improve the efficiency of the parameter learning process by an order of magnitude, while leading to more accurate link prediction models in a significantly lower number of iterations. We extensively evaluate the proposed learning procedure on a variety of new models: our result show a significant improvement over state-of-the-art link prediction methods on two large Knowledge Graphs, namely WordNet and Freebase. [ABSTRACT FROM AUTHOR]

In the context of semantic knowledge bases, among the possible problems that may be tackled by means of data-driven inductive strategies, one can consider those that require the prediction of the unknown values of existing numeric features or the definition of new features to be derived from the data model. These problems can be cast as regression problems so that suitable solutions can be devised based on those found for multi-relational databases. In this paper, a new framework for the induction of logical regression trees is presented. Differently from the classic logical regression trees and the recent fork of the terminological classification trees, the novel terminological regression trees aim at predicting continuous values, while tests at the tree nodes are expressed with Description Logic concepts. They are intended for multiple uses with knowledge bases expressed in the standard ontology languages for the Semantic Web. A top-down method for growing such trees is proposed as well as algorithms for making predictions with the trees and deriving rules. The system that implements these methods is experimentally evaluated on ontologies selected from popular repositories. [ABSTRACT FROM AUTHOR]