Contrastive learning of adverse events to provide effective and interpretable vector representations for machine-assisted pharmacovigilance

Balogh Olivér Márton; Pétervári Mátyás; Csernák Áron Márk; Puhl Eszter; Horváth András; Ferdinandy Péter; Ágg Bence: Contrastive learning of adverse events to provide effective and interpretable vector representations for machine-assisted pharmacovigilance.
BRIEFINGS IN BIOINFORMATICS, 27 (5). ISSN 1467-5463 (2026)

Szöveg
bbag463.pdf - Megjelent verzió
Download (4MB)

Mű típusa:

Folyóiratcikk

Szerző azonosítók:

Név	ORCID	MTMT szerző azonosító
Balogh Olivér Márton	0000-0002-6296-6383	10082699
Pétervári Mátyás	0000-0002-4816-0746	10065329
Csernák Áron Márk	0000-0002-6040-4619	10100490
Puhl Eszter	0000-0001-8501-1478	10072501
Horváth András	0000-0001-5855-4186	10029872
Ferdinandy Péter	0000-0002-6424-6806	10010125
Ágg Bence	0000-0002-6492-0426	10057401

Absztrakt (kivonat):

Post-marketing surveillance is crucial for drug safety, yet the tools of pharmacovigilance rely solely on text-based data that may limit contemporary machine learning methodologies in the support of decision-making. With the recent surge of employing large language models (LLMs) for text-based tasks, there also arises an unmet need for a different approach which is not grounded in the linguistic patterns of unfiltered natural text, like LLMs, but rather based on real-world drug safety data. Here, we adapt contrastive learning algorithms to generate adverse event vector representations from spontaneous adverse event reports to serve as machine-readable (i.e. numerical) resources for downstream pharmacovigilance applications, such as drug-event association prediction for signal detection or causality assessment. We present comprehensive interpretability analyses of the resulting representations through density-based clustering, semantic evaluation, and comparison of multivariate dispersions, revealing patterns that reflect both functional and causal relations of the adverse events while also capturing drug-safety-related information better than existing medical terminologies and encoder-only LLMs. Furthermore, we demonstrate the applicability of our representations as input features in our downstream classifier model, outperforming the reporting odds ratio method, commonly used by regulatory agencies, and also LLM-generated representations (area under the receiver operating characteristic curve: 0.88 versus 0.76-0.83) on drug-event association prediction benchmarks. Therefore, we propose an interpretable adverse event vector representation, serving as a general resource that could enable the development of a wide array of machine learning applications to support decision-making in pharmacovigilance and facilitate patient safety.

Folyóirat címe:

BRIEFINGS IN BIOINFORMATICS

Megjelenés éve:

2026

Kötet:

Szám:

ISSN:

1467-5463

Intézmény:

Pázmány Péter Katolikus Egyetem

Kar:

Információs Technológiai és Bionikai Kar (2013.07.-)