Reconstructing Biological Pathways by Applying Selective Incremental Learning to (Very) Small Language Models

TL;DR

This study demonstrates that a small, domain-specific language model can accurately predict biological pathway interactions, improving understanding of intracellular processes with minimal data and active learning strategies.

Contribution

The paper introduces a novel approach using a very small BERT-based model with active learning to reconstruct biological pathways, reducing data requirements and enhancing accuracy.

Findings

Achieved over 80% accuracy in predicting molecular interactions.

Active learning with entropy-based selection improves model performance.

Small models can effectively contribute to biomedical pathway reconstruction.

Abstract

The use of generative artificial intelligence (AI) models is becoming ubiquitous in many fields. Though progress continues to be made, general purpose large language AI models (LLM) show a tendency to deliver creative answers, often called "hallucinations", which have slowed their application in the medical and biomedical fields where accuracy is paramount. We propose that the design and use of much smaller, domain and even task-specific LM may be a more rational and appropriate use of this technology in biomedical research. In this work we apply a very small LM by today's standards to the specialized task of predicting regulatory interactions between molecular components to fill gaps in our current understanding of intracellular pathways. Toward this we attempt to correctly posit known pathway-informed interactions recovered from manually curated pathway databases by selecting and using only the most informative examples as part of an active learning scheme. With this example we show that a small (~110 million parameters) LM based on a Bidirectional Encoder Representations from Transformers (BERT) architecture can propose molecular interactions relevant to tuberculosis persistence and transmission with over 80% accuracy using less than 25% of the ~520 regulatory relationships in question. Using information entropy as a metric for the iterative selection of new tuning examples, we also find that increased accuracy is driven by favoring the use of the incorrectly assigned statements with the highest certainty (lowest entropy). In contrast, the concurrent use of correct but least certain examples contributed little and may have even been detrimental to the learning rate.