ellimilial/clinical-trials-negative-efficacy-biomedbert

Negative Efficacy Classifier (BiomedBERT)

Base model: microsoft/BiomedNLP-BiomedBERT-base-uncased-abstract-fulltext

This model was fine-tuned to classify whether a clinical trial termination text indicates negative efficacy / futility / unmet endpoint.

Labels

• —0 = notnegativeefficacy
• —1 = negative_efficacy

Training data

Fine-tuned on the Open Targets dataset opentargets/clinical_trial_stop_reasons, introduced in Razuvayevskaya, O., Lopez, I., Dunham, I. et al. Genetic factors associated with reasons for clinical trial stoppage. Nature Genetics 56, 1862–1867 (2024). https://doi.org/10.1038/s41588-024-01854-z

Inference note

A suggested operating threshold is stored in inference_config.json.

Text normalisation used

def normalise_text(text):
    text = "" if pd.isna(text) else str(text)
    text = text.replace("&", "&")
    text = re.sub(r"\s+", " ", text.strip().lower())
    return text