Biological rationales from language models enable leakage-resistant forecasts of target-indication success
- 1UT MD Anderson Cancer Center
Abstract
Identifying which target-indication (T-I) hypotheses can translate into clinical success remains a central challenge in drug discovery. We present PRIORITI (Prospective Rationale-Informed Outcome Reasoning for Integrated Target-indication Intelligence), a leakage-resistant framework that separates upstream biological evidence synthesis from outcome prediction. Given only a target gene and disease indication, a domain-instructed large language model (LLM) synthesizes drug-agnostic human genetics across prespecified channels. These evidence were embedded and combined with static entity context derived from gene summaries and disease definitions. The resulting representation is used to fit a supervised model trained on 6,784 temporally curated historical T-I outcomes, which is one of the largest cohorts assembled for this task. On a held-out historical test set of 1,696 T-I pairs, PRIORITI achieved a ROC-AUC of 0.883 and an expected calibration error (ECE) of 0.023, outperforming both a static target and indication context baseline and a human-curated genetic-evidence reference. On a strictly post-cutoff out-of-time benchmark of 765 T-I pairs, the largest such evaluation cohort reported to date, performance remained robust, with ROC-AUC 0.817, PR-AUC 0.357 and ECE 0.044, exceeding static target and indication context baseline and direct zero-shot LLM forecasting. A label-blinded LLM-based explainer further converted locked predictions into auditable prioritization rationales. Together, PRIORITI provides a calibrated, biology-first probability together with decision-oriented rationales for T-I prioritization before downstream translational investment.
Related
- A Context-Aware 14-3-3 Binding Predictor Enabled by Data Augmentation, Multi-Scale Biological Features, and Protein Language Model Embeddings
- Biological and genomic characterization of Escherichia phage KBU_SV1 active against carbapenem-resistant Escherichia coli
- The phage stress test: a proving ground for genome language models in biological prediction
- Structured biological evidence enables general-purpose language models to predict bacteriophage-host interactions
- Morphological and biochemical characterization of Capsicum annuum landrace Durgada for potential geographical indication status in Andhra Pradesh, India
BibTeX
@misc{zhang2026biological,
title = {{Biological rationales from language models enable leakage-resistant forecasts of target-indication success}},
author = {W. Zhang and J. Xu and Z. Zhang and M. Liu and R. Sun and B. Al-lazikani and X. Shen and S. Kopetz and L. Wu and B. Zhao and C. Wu},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.24.754137},
url = {https://www.biorxiv.org/content/10.64898/2026.09.24.754137v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Biological rationales from language models enable leakage-resistant forecasts of target-indication success
AU - W. Zhang
AU - J. Xu
AU - Z. Zhang
AU - M. Liu
AU - R. Sun
AU - B. Al-lazikani
AU - X. Shen
AU - S. Kopetz
AU - L. Wu
AU - B. Zhao
AU - C. Wu
PY - 2026
DA - 2026/09/24
PB - bioRxiv
DO - 10.64898/2026.09.24.754137
UR - https://www.biorxiv.org/content/10.64898/2026.09.24.754137v1
AB - Identifying which target-indication (T-I) hypotheses can translate into clinical success remains a central challenge in drug discovery. We present PRIORITI (Prospective Rationale-Informed Outcome Reasoning for Integrated Target-indication Intelligence), a leakage-resistant framework that separates upstream biological evidence synthesis from outcome prediction. Given only a target gene and disease indication, a domain-instructed large language model (LLM) synthesizes drug-agnostic human genetics across prespecified channels. These evidence were embedded and combined with static entity context derived from gene summaries and disease definitions. The resulting representation is used to fit a supervised model trained on 6,784 temporally curated historical T-I outcomes, which is one of the largest cohorts assembled for this task. On a held-out historical test set of 1,696 T-I pairs, PRIORITI achieved a ROC-AUC of 0.883 and an expected calibration error (ECE) of 0.023, outperforming both a static target and indication context baseline and a human-curated genetic-evidence reference. On a strictly post-cutoff out-of-time benchmark of 765 T-I pairs, the largest such evaluation cohort reported to date, performance remained robust, with ROC-AUC 0.817, PR-AUC 0.357 and ECE 0.044, exceeding static target and indication context baseline and direct zero-shot LLM forecasting. A label-blinded LLM-based explainer further converted locked predictions into auditable prioritization rationales. Together, PRIORITI provides a calibrated, biology-first probability together with decision-oriented rationales for T-I prioritization before downstream translational investment.
ER -