LLM-Augmented Computational Phenotyping of Long Covid

2026-03-18

AI Summary

This study addresses the high heterogeneity and lack of systematic phenotypic characterization in long COVID clinical subtypes, which hinders precision interventions. The authors propose “Grace Cycle,” a computational phenotyping framework that, for the first time, integrates large language models into a statistically rigorous longitudinal data analysis pipeline. Through an iterative mechanism of hypothesis generation, evidence extraction, and feature refinement, the framework enables interpretable, disease-agnostic subtype discovery. Applied to data from 13,511 patients, the method identifies three distinct subtypes—Protected, Responder, and Refractory—that exhibit significant differences in symptom severity, baseline disease burden, and patterns of dose–response to treatment. These findings establish a novel paradigm for precise subtyping and targeted intervention in long COVID.

Technology Category

Natural Language Processing: Language Grounding & Multi-modal NLPMachine Learning: Large Multimodal Models (LMMs)Application Domains: Humanities & Computational Social Science

Application Category

Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphsUser Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systemsEconomics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Phenotypic characterization is essential for understanding heterogeneity in chronic diseases and for guiding personalized interventions. Long COVID, a complex and persistent condition, yet its clinical subphenotypes remain poorly understood. In this work, we propose an LLM-augmented computational phenotyping framework ``Grace Cycle'' that iteratively integrates hypothesis generation, evidence extraction, and feature refinement to discover clinically meaningful subgroups from longitudinal patient data. The framework identifies three distinct clinical phenotypes, Protected, Responder, and Refractory, based on 13,511 Long Covid participants. These phenotypes exhibit pronounced separation in peak symptom severity, baseline disease burden, and longitudinal dose-response patterns, with strong statistical support across multiple independent dimensions. This study illustrates how large language models can be integrated into a principled, statistically grounded pipeline for phenotypic screening from complex longitudinal data. Note that the proposed framework is disease-agnostic and offers a general approach for discovering clinically interpretable subphenotypes.

Problem

• Long COVID
• computational phenotyping
• clinical subphenotypes
• disease heterogeneity
• personalized interventions

Innovation

• LLM-augmented phenotyping
• computational phenotyping
• longitudinal data analysis
• clinical subphenotypes
• Grace Cycle