KGRD: a knowledge-graph-augmented automated reasoning framework for diagnosis and counselling of paediatric rare genetic disorders
6 affiliations
- 1National Institute for Data Science in Health and Medicine, Xiamen University, Xiamen, China
- 2Institute of Artificial Intelligence, Xiamen University, Xiamen, China
- 3School of Medicine, Xiamen University, Xiamen, China
- 4School of Pharmaceutical Sciences, Xiamen University, Xiamen, China
- 5Kindstar Globalgene Technology, Inc., Wuhan, China
- 6Department of Paediatrics, Women and Children’s Hospital, School of Medicine, Xiamen University, Xiamen, China
Abstract
Diagnosis and counselling for paediatric rare diseases remain constrained by the sparsity of structured patient-level data and fragmented genetic knowledge, which can induce ‘common-attention’ bias in conventional large language models (LLMs). Here, we describe KGRD, a knowledge-graph-augmented diagnostic support framework empowered by knowledge-driven and data-driven inference over patient-level genomic and phenotypic data. KGRD consists of three specialised inference agents for deductive reasoning regarding disease aetiology, as well as a collective decision-making module that integrates multidisciplinary deliberation with multi-source verification. In a validation benchmark of 420 rare disease cases, KGRD(DS) achieved the strongest overall performance, increased the mean Bond score of the top-ranked diagnosis from 3.27 to 3.85 and raised CIE from 73.6 to 81.9%, corresponding to 35 additional cases with candidate-diagnosis Bond score ≥4. Together, these results indicate that KGRD provides effective diagnostic support for paediatric rare diseases.
Related
- Temporal-spatial reasoning over hypergraph knowledge structures for multimodal retrieval-augmented generation
- Fertility preservation counselling in women with cancer: knowledge, preferences, and unmet needs from a cross-sectional survey.
- Knowledge, attitudes, and practices on paediatric tuberculosis in rural Tanzania: insights for enhanced detection and care.
- Integrating knowledge graphs and retrieval-augmented generation to identify clinical pharmacology gaps in infectious disease trial protocols
- HalluciGuard-eye: knowledge graph-enhanced retrieval-augmented generation for hallucination detection and mitigation in ophthalmic multimodal large language models
BibTeX
@article{guo2026kgrd,
title = {{KGRD: a knowledge-graph-augmented automated reasoning framework for diagnosis and counselling of paediatric rare genetic disorders}},
author = {Guiling Guo and Zhenyu Shao and Heying Luo and Ziqing Fu and Hui Xiong and Jin Li and Qingqing Qin and Xiaoqing Yang and Shuxiang Hu and Jinzhun Wu and Qiyuan Li},
journal = {NPJ Digital Medicine},
year = {2026},
month = jul,
doi = {10.1038/s41746-026-02943-5},
url = {https://doi.org/10.1038/s41746-026-02943-5},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - JOUR
TI - KGRD: a knowledge-graph-augmented automated reasoning framework for diagnosis and counselling of paediatric rare genetic disorders
AU - Guiling Guo
AU - Zhenyu Shao
AU - Heying Luo
AU - Ziqing Fu
AU - Hui Xiong
AU - Jin Li
AU - Qingqing Qin
AU - Xiaoqing Yang
AU - Shuxiang Hu
AU - Jinzhun Wu
AU - Qiyuan Li
PY - 2026
DA - 2026/07/02
JO - NPJ Digital Medicine
DO - 10.1038/s41746-026-02943-5
UR - https://doi.org/10.1038/s41746-026-02943-5
AB - Diagnosis and counselling for paediatric rare diseases remain constrained by the sparsity of structured patient-level data and fragmented genetic knowledge, which can induce ‘common-attention’ bias in conventional large language models (LLMs). Here, we describe KGRD, a knowledge-graph-augmented diagnostic support framework empowered by knowledge-driven and data-driven inference over patient-level genomic and phenotypic data. KGRD consists of three specialised inference agents for deductive reasoning regarding disease aetiology, as well as a collective decision-making module that integrates multidisciplinary deliberation with multi-source verification. In a validation benchmark of 420 rare disease cases, KGRD(DS) achieved the strongest overall performance, increased the mean Bond score of the top-ranked diagnosis from 3.27 to 3.85 and raised CIE from 73.6 to 81.9%, corresponding to 35 additional cases with candidate-diagnosis Bond score ≥4. Together, these results indicate that KGRD provides effective diagnostic support for paediatric rare diseases.
SN - 2398-6352
ER -