AI identifies patterns in biomedical findings to suggest future discoveries
Learning Scientific Trajectories to Predict Future Biomedical Discoveries
Computation and LanguageArtificial Intelligence
Summary
Finding new ideas in biomedical research is hard because of the huge amount of information published over time. The authors created a system called ForeSCITE that looks at how scientific conclusions change over time in a special language space and uses this to suggest new possible discoveries. Their method can predict new biomedical ideas that are not yet explored but seem plausible based on past trends and expert checks. This approach helps automate hypothesis generation for future scientific studies.
What this means in practice
- •For biomedical data teams: Generate novel biomedical hypotheses by analyzing temporal patterns in published scientific conclusions to guide experimental directions.
- •For pharmaceutical research teams: Prioritize promising therapeutic hypotheses that are novel and biologically plausible by applying AI-driven temporal semantic analysis to scientific literature.
Authors
Chia-Hsuan Chang, Brian Ondov, Jiang Lv, Lan Jiang, Yunmai Wang, Haerin Song, Xueqing Peng, Qiaozhu Mei, Xiaolei Su, Hua Xu
Abstract
AI co-scientists have expanded the use of large language models for generating scientific hypotheses and guiding experimental investigation. Here we investigate whether temporal patterns in the semantic representations of complete biomedical conclusions can guide hypothesis generation. We introduce ForeSCITE, a framework that identifies thematic clusters of published conclusions, estimates a direction separating earlier from more recent findings within each cluster, and projects existing conclusion embeddings along these directions. An embedding language model trained to reconstruct conclusions decodes the projected coordinates into natural-language hypotheses. Across 407 clusters derived from approximately three million pre-cutoff conclusions, we identify temporal structure and illustrate its retention after decoding. In retrospective, time-sliced evaluations, ForeSCITE increases semantic coverage beyond the historical corpus, biomedical entity-pair novelty and post-cutoff pair re-observation relative to five prompt-based and retrieval-augmented baselines. Expert review of prioritized CAR-T hypotheses identified 37 candidates as both novel and biologically plausible, and expert justifications for five selected hypotheses illustrate their biological rationale and subsequent literature support where available. These findings support temporal guidance in semantic space as a scalable mechanism for proposing biomedical hypotheses for subsequent scientific evaluation.