MSACL 2025 Abstract

Development and Application of a New Assay Validation Tool on a Series of Cardiovascular Disease-Related Peptide Biomarkers

Dan Lane (1, 2), Borislav Lazarov (1, 3), Colleen Maxwell (1, 2), Leong L Ng (1, 2), Donald JL Jones (1, 2, 4), Pankaj Gupta (1, 3)
(1) Department of Cardiovascular Sciences, University of Leicester, Leicester, UK, (2) van Geest MS-OMICS facility, University of Leicester, Leicester, UK, (3) The Department of Chemical Pathology and Metabolic Diseases, Leicester Royal Infirmary, University Hospitals of Leicester, UK, (4) Leicester Cancer Research Centre and Department of Genetics and Genome Biology, University of Leicester, Leicester, UK

Dan Lane (Presenter) University of Leicester

Presenter Bio: I am a post-doctoral researcher working with translational mass spectrometry in the cardiovascular disease and infectious diseases.

Relevant Financial Disclosures (within past 24 months, reported on Mar 18, 2025) No relevant financial relationship(s) to disclose.

INTRODUCTION:
Mass spectrometry assay validation is essential for ensuring accuracy, precision, and robustness in clinical applications. However, current processes are slow and highly heterogeneous across laboratories. Validation often takes weeks per assay due to the burdensome manipulation of output batch files in Excel and the preparation of accreditation-reports. Many putative protein biomarkers with potential clinical value are reported each year from discovery studies, yet the slow development and validation of clinically viable assays creates a major bottleneck in translating research into clinical practice. To address this challenge, an interactive and automated tool was developed to process mass spectrometric validation data and generate accreditation-ready reports. We trialed this tool on the in-house validation of several peptide biomarkers for cardiovascular diseases.

OBJECTIVE(S):
We sought to develop a vendor-agnostic tool that streamlines and automates mass spectrometry assay validation. By integrating data processing, statistical analysis, and report generation features into a single platform, our goal was to significantly reduce the time and cost associated with validation while maintaining compliance with accreditation standards. We sought to test the tools’ efficacy against a peptide-level quantification assay that leveraged the automated AssayMap workflow on the Agilent Bravo.

METHODS:
The tool - Validation Processor - was developed using the R programming language. We designed an interactive graphical user interface that facilitated user-friendly data analysis. It allowed users to upload data from multiple vendor output files including Skyline (.csv), Waters (.txt), and Shimadzu (.csv). The Validation Processor was built to include response function characterisation for calibration, bias and precision review across batches, analyte stability monitoring, and other critical parameters. The software included functionality to set acceptance criteria in accordance with established standards such as ANSI/ASB Standard 036 for Method Validation. A report generator script compiled the data into an accreditation-ready document that summarised results, highlighted peptides that passed or failed the validation criteria, and incorporated essential system information for traceability.

To evaluate the tool, a quantitative assay for 10 tryptic peptides was developed in plasma for the C-reactive protein, vitamin D-binding protein, and apolipoprotein A-IV biomarkers. Leveraging the automated workflow of the Agilent Bravo, the liquid-handling robot was fitted with the AssayMAP head, and acidified post-digested plasma peptides were cleaned using C18 5 µL cartridges. Plasma was pooled, diluted 100-fold to act as surrogate matrix, and subsequently fortified with recombinant peptide standards along with heavy labels to fulfil the validation experiments. Samples were analysed across 5 batches using a Waters Xevo TQ-XS.

RESULTS:
Peak integration was interrogated in Skyline, and response data were exported as .csv files. These were then imported into the Validation Processor and calibration models, bias, precision, carryover, interference, ionisation suppression, detection and quantification limits, stability, and dilution integrity were all calculated within 30 seconds under ANSI/ASB Standard 036 criteria (run locally: AMD Ryzen 7 4700U, 8 GB RAM).

The automated processor generated individual graphical and tabular outputs for each validation experiment (e.g., calibration curves, boxplots), providing a clear and structured overview of the assay performance. A comprehensive validation report (.docx or .pdf) was generated in under one-minute, significantly reducing the time required for data review and reporting. Compared to manual validation, which can often take weeks, this automated approach could save laboratories upwards of £2000 per validation (3-week staff time equivalent).

In summary of the peptide assay validated using the Validation Processor, the average R2 was 0.997 (SD: 0.003; all processed with linear 1/x weighted models), limit of detection was 0.35 fmol/µL (SD: 0.35), bias was at 2.5% (SD: 7.9), intra-batch precision was 6.3% (SD: 4.8), and inter-batch precision was 3.9% (SD: 2.6). The assay was demonstrably suitable for peptide level quantitation of cardiovascular disease protein biomarkers.

CONCLUSION:
The Validation Processor is a novel tool that standardises and accelerates mass spectrometry assay validation. By significantly reducing the time and manual effort required for data analysis and reporting, it has the potential to improve efficiency in clinical, research, and industrial laboratories. The tool could facilitate faster translation of biomarkers, bridging the gap between discovery research and clinical utility.