Contexts and contradictions: a roadmap for computational drug repurposing with knowledge inference

Abstract

The cost of drug development continues to rise and may be prohibitive in cases of unmet clinical need, particularly for rare diseases. Artificial intelligence-based methods are promising in their potential to discover new treatment options. The task of drug repurposing hypothesis generation is well-posed as a link prediction problem in a knowledge graph (KG) of interacting of drugs, proteins, genes and disease phenotypes. KGs derived from biomedical literature are semantically rich and up-to-date representations of scientific knowledge. Inference methods on scientific KGs can be confounded by unspecified contexts and contradictions. Extracting context enables incorporation of relevant pharmacokinetic and pharmacodynamic detail, such as tissue specificity of interactions. Contradictions in biomedical KGs may arise when contexts are omitted or due to contradicting research claims. In this review, we describe challenges to creating literature-scale representations of pharmacological knowledge and survey current approaches toward incorporating context and resolving contradictions.

Utilizing LLMs for Enhanced Argumentation and Extraction of Causal Knowledge from Scientific Literature

“…This would highly influence the quality of its downstream task. For instance, since the concepts extracted are overly general, it can result in many apparent contradictions that are not truly contradictory (9). Additionally, existing natural language processing (NLP) techniques for biomedical entity and relations extraction frequently neglect the quality of arguments, especially the varying degrees of strength across different sentences where claims are expressed.…”

Editorial: Emerging areas in literature-based discovery

“…Recent advances in LBD are focused on experimenting with state-of-the-art algorithms such as knowledge graphs, embeddings, deep learning, etc. and expanding the practical use cases of LBD (see for example Sosa and Altman, 2022 ; Syafiandini et al, 2022 ; Zhou et al, 2022 ). Notwithstanding these valuable developments, much can still be done to address more fundamental issues in the field.…”