---
title: "When knowledge graphs improve EHR-based prediction: findings across Parkinson’s, MS, and chronic p"
id: "biorxiv-9-understanding-when-and-how-knowledge-graphs-add-value-to-ehr-based-clinical"
canonical_url: "https://medichelpline.com/clinical-feed/biorxiv-9-understanding-when-and-how-knowledge-graphs-add-value-to-ehr-based-clinical"
content_type: "clinical_feed_article"
specialty: "Neurology"
source_name: "bioRxiv (Biomedical Preprints)"
source_url: "https://www.biorxiv.org/content/10.64898/2026.10.04.756605v1?rss=1"
published_at: "2026-10-11T12:00:00.000Z"
evidence_level: "Verified Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# When knowledge graphs improve EHR-based prediction: findings across Parkinson’s, MS, and chronic p
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/biorxiv-9-understanding-when-and-how-knowledge-graphs-add-value-to-ehr-based-clinical
- **Specialty:** [Neurology](https://medichelpline.com/clinical-feed/neurology.md)
- **Primary Source:** bioRxiv (Biomedical Preprints)
- **Source URL:** [Original Journal Publication](https://www.biorxiv.org/content/10.64898/2026.10.04.756605v1?rss=1)
- **Published At:** 2026-10-11T12:00:00.000Z
- **Evidence Rating:** Verified Feed
## Executive GIST (TL;DR)
- The authors evaluated the additive value of **knowledge graphs (KGs)** for machine-learning prediction using electronic health record (**EHR**) data across three diseases: **Parkinson’s disease**, **multiple sclerosis**, and **chronic pain**. - Five modelling strategies were compared for each disease: EHR-only features, and four approaches incorporating graph diffusion over different SPOKE-derived subgraphs: concept-level, cohort-derived, disease-anchored, and the full SPOKE graph. - **Cohort-derived subgraphs** contained roughly 45% of SPOKE nodes and about 20% of edges; they matched or slightly outperformed the full graph while requiring substantially less computation across all three diseases. - The predictive impact of KG integration was disease-dependent: it improved classification performance in Parkinson’s disease, had no measurable effect in multiple sclerosis, and decreased performance in chronic pain. - Differences in benefit were not explained by the amount of molecular or biological information incorporated: the disease with the largest shift toward molecular features did not improve, the disease that benefitted showed only a modest molecular shift, and the disease with no molecular shift was harmed. - Mechanistically, **graph diffusion** redistributed existing signal across many correlated features rather than introducing independent biological signal; hence performance change depended on whether that redistribution added information beyond EHR inputs. - The study suggests cohort-derived subgraphs as a computationally efficient alternative to full-scale KGs and frames KG integration as a **data- and disease-dependent transformation** of patient representation rather than a universal accuracy booster. - The authors report no competing interests.
## Clinical Analysis & Structured Key Points
Understanding when and how knowledge graphs add value to EHR-based clinical prediction across Parkinson's disease, multiple sclerosis, and chronic pain | bioRxiv Skip to main content New Results Understanding when and how knowledge graphs add value to EHR-based clinical prediction across Parkinson's disease, multiple sclerosis, and chronic pain Aishwarya Pathak , Oleg Stroganov , Tracy Ballinger , Tania Khasanova doi: https://doi.org/10.64898/2026.10.04.756605 Aishwarya Pathak Rancho BioSciences Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: aish.pathak24{at}gmail.com Oleg Stroganov Rancho BioSciences Find this author on Google Scholar Find this author on PubMed Search for this author on this site Tracy Ballinger Rancho BioSciences Find this author on Google Scholar Find this author on PubMed Search for this author on this site Tania Khasanova Rancho BioSciences Find this author on Google Scholar Find this author on PubMed Search for this author on this site Abstract Info/History Metrics Supplementary material Preview PDF Abstract Biomedical knowledge graphs (KGs) have emerged as a promising approach for incorporating biological knowledge into machine-learning models trained on electronic health record (EHR) data. However, their large scale and heterogeneous structure impose substantial computational cost, and it remains unclear whether their benefits generalize across diseases or depend on how the graph is constructed. We systematically evaluated KG integration across three diseases spanning distinct biological and clinical characteristics: multiple sclerosis, Parkinson's disease, and chronic pain. For each disease we compared five modelling strategies, spanning EHR-only features and graph diffusion over concept-level, cohort-derived, disease-anchored, and full SPOKE (Scalable Precision Medicine Open Knowledge Engine) subgraphs, asking whether graph construction affects predictive performance and computational cost, whether KG integration improves classification over EHR-only models, and what the graph does to model behavior. Cohort-derived subgraphs, comprising roughly 45% of SPOKE's nodes and one-fifth of its edges, matched or modestly exceeded the full graph's performance while substantially reducing computational cost in all three diseases. The predictive benefit of KG integration, however, was strongly disease-dependent: it improved classification in Parkinson's disease, had no measurable effect in multiple sclerosis, and reduced performance in chronic pain. These differences were not explained by the amount of biological information incorporated: the disease with the largest shift toward molecular features showed no improvement, the sole beneficiary showed only a modest shift, and the disease with no molecular shift was penalized. Mechanistically, graph diffusion redistributed each disease's existing signal across many correlated features rather than adding independent biological signal, so that any gain depended on whether it added information beyond the EHR. These findings show that cohort-derived subgraphs offer a computationally efficient alternative to full-scale knowledge graphs, and that KG integration functions less as a universal accuracy booster than as a data- and disease-dependent transformation of the patient representation. Competing Interest Statement The authors have declared no competing interest. Copyright The copyright holder for this preprint is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY-NC-ND 4.0 International license . Back to top Previous Next Posted October 11, 2026. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Understanding when and how knowledge graphs add value to EHR-based clinical prediction across Parkinson's disease, multiple sclerosis, and chronic pain Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Understanding when and how knowledge graphs add value to EHR-based clinical prediction across Parkinson's disease, multiple sclerosis, and chronic pain Aishwarya Pathak , Oleg Stroganov , Tracy Ballinger , Tania Khasanova bioRxiv 2026.10.04.756605; doi: https://doi.org/10.64898/2026.10.04.756605 Share This Article: Copy Citation Tools Understanding when and how knowledge graphs add value to EHR-based clinical prediction across Parkinson's disease, multiple sclerosis, and chronic pain Aishwarya Pathak , Oleg Stroganov , Tracy Ballinger , Tania Khasanova bioRxiv 2026.10.04.756605; doi: https://doi.org/10.64898/2026.10.04.756605 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Areas All Articles Animal Behavior and Cognition (8110) Biochemistry (18940) Bioengineering (15092) Bioinformatics (44968) Biophysics (22850) Cancer Biology (19967) Cell Biology (27178) Clinical Trials (138) Developmental Biology (14077) Ecology (21277) Epidemiology (2067) Evolutionary Biology (25676) Genetics (16278) Genomics (23726) Immunology (18903) Microbiology (42964) Molecular Biology (18237) Neuroscience (94490) Paleontology (709) Pathology (3014) Pharmacology and Toxicology (5149) Physiology (8217) Plant Biology (16165) Scientific Communication and Education (2106) Synthetic Biology (4610) Systems Biology (10331) Zoology (2408)
## Related Clinical Research

- [Mitochondrial dysfunction in neurology: biomarkers for diagnosis, stratification and monitoring](https://medichelpline.com/clinical-feed/nature-reviews-neurology-0-mitochondrial-dysfunction-in-neurology-biomarkers-and-clinical-implications.md)
- [Exercise Interventions Improve Gait, Motor Function, and Some Balance Measures in Parkinson’s Dise](https://medichelpline.com/clinical-feed/plos-one-14-effects-of-exercise-interventions-on-gait-motor-function-and-balance-in.md)
- [Ultra-processed foods linked to MRI signs of vascular brain injury in middle-aged adults](https://medichelpline.com/clinical-feed/medical-news-today-0-ultra-processed-foods-may-be-harming-brain-health-in-otherwise-healthy-adults.md)
- [Global dementia incidence projected to reach 24.7 million new cases annually by 2050](https://medichelpline.com/clinical-feed/medical-news-today-0-dementia-cases-to-double-by-2050-why-women-some-countries-may-be-at-higher-risk.md)
- [Effector-memory T cell program links peripheral EBV priming to CNS autoimmunity — preprint summary](https://medichelpline.com/clinical-feed/biorxiv-1-a-pro-inflammatory-effector-memory-t-cell-program-links-peripheral-ebv-priming.md)

## Navigation
- [← Back to Neurology Feed](https://medichelpline.com/clinical-feed/neurology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.