Patient-specific Biomolecular Instruction Tuning

2025-09-26

AI Summary

Current multimodal large language models (MLLMs) struggle to perform clinical-grade molecular interpretation of patient-specific proteomic data, primarily due to the absence of oncology-focused proteomic instruction-tuning datasets and the inability of existing architectures to capture topological heterogeneity in molecular interactions. To address this, we introduce CPTAC-PROTSTRUCT—the first oncology-specific proteomic instruction dataset—and propose KRONOS, a graph-enhanced multimodal LLM framework that jointly integrates proteomic abundance profiles, molecular interaction network topology, and biomedical knowledge graphs, enabling end-to-end clinical reasoning via instruction tuning. Evaluated on molecular subtype classification, longitudinal trajectory modeling, and tumor staging prediction, KRONOS significantly outperforms baseline models. It achieves, for the first time, interpretable and generalizable mapping from proteomic data to personalized clinical decision-making—including diagnosis, prognosis, and therapeutic stratification—thereby bridging a critical gap between deep proteomics and precision oncology.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Data Mining & Knowledge Management: Representing, Reasoning, and Using Provenance, Trust

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation

Abstract

Proteomics data is essential to pathogenic understanding of a disease phenotype. In cancer, analysis of molecular signatures enables precision medicine through the identification of biological processes that drive individualized tumor progression, therapeutic resistance, and clinical heterogeneity. Recent advances in multimodal large language models (LLMs) have shown remarkable capacity to integrate and reason across heterogeneous data modalities. However, performing multi-modal language modeling for molecular understanding of patient-specific proteomics remains a significant challenge due to two barriers: (1) the lack of instruction-tuning datasets that enable clinical interpretation from proteomics data, and (2) the absence of language modeling architectures designed to capture the rich heterogeneity of molecular data. In this work, we introduce CPTAC-PROTSTRUCT, the first instruction tuning dataset for molecular understanding of oncology, comprising over 400k open-ended examples derived from individualized proteomic profiles curated from the largest national proteomics cancer study (CPTAC). Additionally, we propose KRONOS (Knowledge Representation of patient Omics Networks in Oncology via Structured tuning), a novel graph-LLM framework that leverages molecular interaction topology with proteomics to learn patient-specific graph representations for enhanced clinical reasoning. We show that KRONOS achieves competitive performance across benchmark clinical tasks, including molecular classification, temporal trajectory modeling, and tumor stage prediction from proteomics data. Ultimately, this approach empowers LLMs to understand patient-level pathogenesis, advancing precision medicine through more accurate diagnosis, prognosis, and treatment stratification.

Problem

• Lack of instruction-tuning datasets for clinical interpretation from proteomics data
• Absence of language modeling architectures for molecular data heterogeneity
• Challenges in patient-specific proteomics understanding using multimodal language models

Innovation

• Introduces CPTAC-PROTSTRUCT instruction dataset for oncology proteomics
• Proposes KRONOS graph-LLM framework using molecular interaction topology
• Learns patient-specific graph representations for clinical reasoning tasks