# Enhanced Diagnosis of Antibody-Mediated Kidney Rejection by Machine Learning and Hybrid Targeted-Shotgun Proteomics

> **NIH NIH R21** · UNIVERSITY OF PITTSBURGH AT PITTSBURGH · 2020 · $234,008

## Abstract

Kidney transplantation offers the best quality of life for patients with chronic kidney failure. Antibody and T-cell
mediated rejection (ABMR and TCMR) are key factors that determine graft survival. Currently, the diagnosis
and differential diagnosis of rejection relies on histopathologic examination which has known limitations such
as subjective interpretations, limited reproducibility, and the need for expert transplant pathologists. There is an
unmet need to develop more specific and quantitative molecular tests that can complement and enhance
conventional histologic assessment. Among the molecular assays, proteome profiling is more attractive than
genomic and transcriptomic profiling which are subjected to numerous post-translational and epigenetic
regulatory mechanisms. Moreover, morphologic changes form the basis of classifying different allograft
diseases. Therefore, the transplant community needs to invest in biopsy-based assays in addition to the
blood/urine-based assays that are being developed by others. This study is aimed to fully map the proteomic
changes in routinely processed formalin fixed paraffin embedded (FFPE) biopsies using a liquid
chromatography–tandem mass spectrometry (LC-MS/MS) platform. To meet the needs of personalized
medicine, this platform uses a novel strategy and machine learning to simultaneously measure the absolute
expression levels of a panel of targeted biomarkers as well as thousands of untargeted proteins. The central
hypothesis is that LC-MS/MS can be used to define disease-specific biomarkers using a discovery data set,
which can then be followed up by a validation data set to determine if LC-MS/MS based tests can be
implemented in clinical practice. In the current project, we will focus on developing molecular assays for ABMR
since antibody contributes to graft loss in 60% of patients. Two Aims are proposed. In Aim #1, quantitative
proteomic strategies will be used to map proteome-level changes in a discovery set of biopsies with ABMR and
its mimics, such as acute tubular injury (ATI), TCMR, BK virus nephropathy (BKVN), interstitial fibrosis/tubular
atrophy (IFTA), and stable renal function (STA). The goal is to identify potential protein biomarkers that can
distinguish ABMR from its mimics. In Aim #2, the potential ABMR biomarkers obtained in Aim #1 will be
validated and optimized in an independent validation data set. Using a hybrid proteomic platform combing
targeted, shotgun proteomics and machine learning, information on absolute quantitation of potential protein
biomarkers and thousands of other proteins will be collected to build a kidney transplant Protein Atlas for assay
development. Successful completion of this study has great potential to be translated into clinical tests that will
enhance the diagnosis of ABMR from other diseases that can mimic that pathology. It will also serve as a
model for developing a new generation of clincal diagnositc tests that will use routinely fixed biopsy mater...

## Key facts

- **NIH application ID:** 10055012
- **Project number:** 1R21AI148776-01A1
- **Recipient organization:** UNIVERSITY OF PITTSBURGH AT PITTSBURGH
- **Principal Investigator:** PARMJEET S RANDHAWA
- **Activity code:** R21 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2020
- **Award amount:** $234,008
- **Award type:** 1
- **Project period:** 2020-06-03 → 2021-05-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10055012

## Citation

> US National Institutes of Health, RePORTER application 10055012, Enhanced Diagnosis of Antibody-Mediated Kidney Rejection by Machine Learning and Hybrid Targeted-Shotgun Proteomics (1R21AI148776-01A1). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10055012. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*