# Extraction of molecular signature of HFpEF via a machine learning-empowered proteomic characterization: A study of the BCAA pathway

> **NIH NIH R01** · UNIVERSITY OF CALIFORNIA LOS ANGELES · 2022 · $649,707

## Abstract

PROJECT SUMMARY
Heart failure with preserved ejection fraction (HFpEF), characterized by heart failure symptoms with normal
ejection fraction, is highly prevalent. However, most HFpEF patients do not respond to standard therapy for heart
failure with reduced ejection fraction (HFrEF), and there are no clear and uniform diagnostic criteria to stratify
and differentiate HFpEF from HFrEF. Therefore, there is a pressing unmet need for us to better understand
HFpEF at the molecular and system levels. Unbiased approaches such as machine learning (ML) offer a powerful
means to tease out the molecular signatures of HFpEF in relevant disease models.
The emerging evidence implicates that metabolism and redox homeostasis are two significant disruptions in
cellular processes evidenced by clinical symptoms of HFpEF. Previous studies have identified branched-chain
amino acid (BCAA) catabolic defect as another major metabolic hallmark in heart failure as well as in metabolic
disorders. Moreover, BCAA catabolic defects have been demonstrated to directly impact mitochondrial function
and elevate reactive oxygen species (ROS) production, resulting in oxidative stress-sensitive post-translational
modifications (O-PTMs) that govern protein function and pathways. These exciting discoveries lead to our new
hypothesis that O-PTM-mediated proteome remodeling is a dynamic and pervasive molecular change in
diseased hearts, affecting proteins with central function in cardiac homeostasis and pathophysiology.
To investigate the unique molecular features and pathogenic mechanisms of HFpEF, we highlight a novel HFpEF
mouse model that incorporates both genetic predisposition for obesity/diabetes and pressure-overload, the two
major risk factors for HFpEF, by performing trans-aortic constriction (TAC) in the ob/ob mice. We have also
perfected the experimental tools and data analysis platform to provide O-PTM profiling at the whole-proteome
level in hearts. Accordingly, we have strategically formulated the following aims according to three phenotypic
levels: At the systemic level, Aim 1 will establish and characterize in vivo mouse models of HFpEF vs. HFrEF
by cardiac and mitochondrial function as well as redox status. At the organellar level, Aim 2 will conduct targeted
proteomics profiling of the cardiac mitochondria and extract O-PTM signatures using ML-based methods to
achieve deep phenotyping of HFpEF and HFrEF. This information will then be integrated and enriched in an O-
PTM molecular atlas and knowledge graph. At the molecular level, Aim 3 will target the BCAA catabolic pathway
to exhaustively scrutinize its role in HFpEF and HFrEF. A multilevel understanding of the HFpEF phenotype,
from its global profiling to molecular targets, will provide valuable new insights into the disease process that can
lead to potential novel diagnostic and therapeutic targets.

## Key facts

- **NIH application ID:** 10440446
- **Project number:** 5R01HL146739-04
- **Recipient organization:** UNIVERSITY OF CALIFORNIA LOS ANGELES
- **Principal Investigator:** Chun Ming Dominic Ng
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2022
- **Award amount:** $649,707
- **Award type:** 5
- **Project period:** 2019-07-01 → 2025-06-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10440446

## Citation

> US National Institutes of Health, RePORTER application 10440446, Extraction of molecular signature of HFpEF via a machine learning-empowered proteomic characterization: A study of the BCAA pathway (5R01HL146739-04). Retrieved via AI Analytics 2026-05-23 from https://api.ai-analytics.org/grant/nih/10440446. Licensed CC0.

---

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