Evaluation of a large-scale aptamer proteomics platform among patients with kidney failure on dialysis

Yue Ren, Peifeng Ruan, Mark Segal, Mirela Dobre, Jeffrey R. Schelling, Upasana Banerjee, Tariq Shafi, Peter Ganz, Ruth F. Dubin

Research output: Contribution to journalArticlepeer-review


Background Patients with kidney failure suffer high mortality, and we currently lack markers for risk stratification for these patients. We carried out a quality control study of a modified aptamer assay (SomaScan v.4.0) that measures ~ 5000 proteins, in preparation for a larger study using this platform in cohorts with kidney failure. Methods Forty participants from the Cardiac, Endothelial Function and Arterial Stiffness in End-Stage Renal Disease (CERES study) were selected to analyze technical and short-term biological variability, orthogonal correlations and differential protein expression in plasma from patients who died during 2.5 year follow-up. Long-term (one year) variability was studied in 421 participants in the Chronic Renal Insufficiency Cohort. We evaluated 4849 aptamers (4607 unique proteins) using data formats including raw data and data formatted using Adaptive Normalization by Maximum Likelihood (ANML), an algorithm developed for SomaScan data in individuals with normal kidney function. Results In ANML format, median[IQR] intra-assay coefficient of variation (CV) was 2.38%[1.76, 3.40] and inter-assay CV was 7.38%[4.61, 13.12]. Short-term within-subject CV was 5.76% [3.35, 9.72]; long-term CV was 8.71%[5.91, 13.37]. Spearman correlations between aptamer and traditional assays for PTH, NT-proBNP, FGF-23 and CRP were all > 0.7. Foldchange (FC) in protein levels among non-survivors, significant after Bonferroni correction, included SVEP1 (FC[95% CI] 2.14 [1.62, 2.82]), keratocan (1.74 [1.40, 2.15]) and LanC-like protein 1 (0.56 [0.45, 0.70]). Compared to raw aptamer data, technical and short-term biological variability in paired samples was lower in ANML-formatted data. ANML formatting had minimal impact on orthogonal correlations with traditional assays or the associations of proteins with the phenotype of mortality. Conclusions SomaScan had excellent technical variability and low within-subject short-term variability. ANML formatting could facilitate comparison of biomarker results with other studies that utilize this format. We expect SomaScan to provide novel and reproducible information in patients with kidney failure on dialysis.

Original languageEnglish (US)
Article numbere0293945
JournalPLoS ONE
Issue number12 December
StatePublished - Dec 2023


  • Humans
  • Renal Dialysis
  • Proteomics/methods
  • Kidney Failure, Chronic/therapy
  • Biomarkers
  • Oligonucleotides

ASJC Scopus subject areas

  • General


Dive into the research topics of 'Evaluation of a large-scale aptamer proteomics platform among patients with kidney failure on dialysis'. Together they form a unique fingerprint.

Cite this