대한신장학회

Skip Navigation: Skip to contents

KOR
ENG
전체메뉴보기

간행물 검색

현재 페이지 경로

HOME
간행물
간행물 검색

논문분류	춘계학술대회 초록집
제목	Large-Scale Systemic Proteome Profiling of Over 5,000 Proteins to Differentiate Primary Glomerular Diseases
저자	Jae-ik Oh
출판정보	2025; 2025(1):
키워드	Primary Glomerular Diseases, Glomerulonephritis, Proteomics, OLINK, Machine learning
초록	Identifying systemic proteome signatures in glomerulonephritis (GN) requires further investigation. Plasma proteome profiling may offer a non-invasive alternative to kidney biopsy for distinguishing GN subtypes. We analyzed plasma samples from GN patients at baseline biopsy, measuring over 5,000 protein levels using the OLINK assay. The cohort included 46 patients with IgA nephropathy (IgAN), 14 with primary focal segmental glomerulosclerosis (FSGS), 48 with minimal change disease (MCD), and 16 with anti-PLA2R antibody-positive membranous nephropathy (MN). The control group had 40 healthy individuals with normal kidney function. We applied Elastic Net regression with 5-fold cross-validation to predict GN subtypes based on proteome profiles with or without baseline clinical features (UPCR, eGFR). To validate the model’s performance, we analyzed plasma proteomes from cohort (32 IgAN, 15 MCD, 14 FSGS, and 10 MN) from other centers. Principal component analysis (PCA) revealed clear discrimination between GN patients and controls. Proteome profile variation was mainly driven by disease category (P = 0.001) rather than proteinuria severity (P = 0.40). A model incorporating 50 plasma proteins achieved strong discriminatory power, yielding an overall ROC-AUC of 0.86 in the validation set. Subtype-specific performance was also strong: MCD (AUC = 0.95), IgAN (AUC = 0.84), and MN (AUC = 0.93). However, performance was relatively lower for FSGS (AUC = 0.69). Notably, including UPCR and eGFR as features in the model did not significantly improve its performance. Top predictive proteins varied by GN subtype. For example, elevated levels of ADAMTS13 and F13B (associated with coagulation) were indicative of MCD, whereas high EFNA4 and SCRIB (related to podocyte injury response) were associated with FSGS. IgAN was characterized by elevated SELL, while MN had high PCOLCE. Distinct plasma proteome signatures exist for each GN subtype, providing a promising tool for non-invasive diagnosis and disease classification.
원문(PDF)	PDF 원문보기

목록

위로가기