The Right Diet Is in Your DNA: Identifying Biomarkers for Personalized Weight Loss

1. Overview

Obesity and overweight are major global health challenges, and personalized nutrition is emerging as a strategy to improve weight-loss outcomes by matching interventions to individual biological profiles. However, clinical responses to commonly prescribed diets, particularly low-carbohydrate versus low-lipid (low-fat) regimens, vary widely among patients, and there is a critical need for predictive biomarkers to guide personalized diet selection. In this pilot project, I will leverage existing transcriptomic and proteomic data from obese or overweight human cohorts to identify molecular signatures that differentiate individuals who respond better to a low-carbohydrate diet versus those who benefit more from a low-lipid diet. By performing integrative re-analysis of these open-access datasets with state-of-the-art genomic and proteomic bioinformatics pipelines, this study aims to uncover genes and proteins whose expression patterns are strongly associated with diet-specific weight loss success. The goal is to lay the groundwork for a future genomics-based diagnostic test that can predict, from a patient’s molecular profile, which dietary strategy will optimize weight loss for that individual. Funding from this award will support data acquisition, computational analysis, and preliminary validation of candidate biomarkers, positioning this project to generate publishable insights and inform larger personalized nutrition studies.

Funding, Facilities, and Ethics Approvals This project relies exclusively on publicly available, de-identified human omics datasets and therefore does not require new human subject recruitment. All data sources comply with prior ethical approvals and open-access sharing policies. Requested funding will support computational infrastructure, dataset acquisition, cloud computing resources, software licenses if needed, and dedicated time for computational analysis and validation.

Estimated Timeline

· Month 1: Identification, curation, and harmonization of suitable open-access transcriptomic and proteomic datasets from diet-intervention cohorts.

· Months 2–4: Integrative bioinformatic analyses to identify candidate gene and protein signatures associated with differential responses to low-carbohydrate versus low-lipid diets.

· Month 5: Cross-dataset validation, pathway enrichment analyses, and prioritization of robust biomarker candidates.

· Month 6: Preparation of manuscript(s), dissemination of findings, and development of a framework for future clinical validation studies.

Data Sharing Commitment All analysis pipelines, processed datasets, and identified biomarker signatures generated during this project will be made openly available through public repositories (e.g., GitHub, Zenodo, or equivalent platforms) in accordance with FAIR data principles. This commitment ensures transparency, reproducibility, and accessibility for the broader research community and supports further advancement of personalized nutrition research.

2. Introduction

Obesity and overweight affect hundreds of millions of people worldwide and are among the strongest modifiable risk factors for type 2 diabetes and cardiovascular disease. Calorie-restricted diets remain the cornerstone of weight-loss recommendations, yet a persistent clinical observation challenges their effectiveness: individuals respond very differently to the same dietary strategy. In particular, both low-carbohydrate and low-fat diets can produce clinically meaningful weight loss over 6–24 months, but average differences between these diets are small, masking substantial interindividual variability in response (Silverii et al., 2022; Volek et al., 2024). This variability has fuelled interest in precision nutrition, the idea that diet efficacy depends on an individual’s biological characteristics. Landmark clinical studies have shown that baseline metabolic traits, such as fasting glucose, insulin levels, and insulin secretion capacity, can predict whether a person loses more weight on a low-carbohydrate versus a low-fat diet (Hjorth et al., 2018; Gardner et al., 2018; Ebbeling et al., 2018). Additional work indicates that specific genetic variants (e.g., APOA2 rs5082), as well as psychosocial and phenotypic factors, interact with diet composition to influence weight-loss trajectories (Lai et al., 2025). These findings collectively suggest that dietary success is not random, but biologically patterned. More recently, multi-omics studies have demonstrated that baseline molecular profiles can predict weight loss and weight regain following calorie-restricted interventions (Li et al., 2025). These advances imply that an individual’s molecular state before dieting encodes predictive information about dietary outcomes. However, while clinical and genetic predictors have been explored, transcriptomic and proteomic signatures that distinguish responders to low-carbohydrate versus low-fat diets remain largely uncharacterized. This represents a critical knowledge gap limiting the translation of precision nutrition into practice. That said, this project is guided by the following questions:

Do baseline gene expression and protein abundance profiles differ between individuals who respond better to low-carbohydrate diets versus those who respond better to low-fat diets?
Can integrative analysis of transcriptomic and proteomic data identify robust molecular signatures predictive of diet-specific weight-loss success?
Are these signatures consistent across independent human cohorts and dietary intervention studies?

Hypothesis Distinct baseline transcriptomic and proteomic signatures in blood or adipose tissue can predict whether an overweight or obese individual will achieve greater weight loss on a low-carbohydrate versus a low-fat calorie-restricted diet, providing predictive power beyond traditional clinical and metabolic markers such as fasting glucose and insulin.

3. Methods

This study will conduct a secondary integrative re-analysis of publicly available randomized trials comparing low-carbohydrate and low-fat calorie-restricted diets in adults with overweight or obesity that include baseline RNA-seq and/or proteomic data and longitudinal weight measurements (≥6–12 months). The primary outcome will be percent weight change at 12 months (or longest follow-up available), and participants will be classified as diet-specific responders if they achieve ≥5% weight loss and lose at least 2 percentage points more than the median of the opposite diet arm. Transcriptomic and proteomic data will be processed using standardized pipelines with quality control, normalization, and batch correction, followed by harmonization of samples and covariates across studies. From these data, gene/protein features and pathway-level scores related to metabolic processes will be generated and integrated with clinical predictors such as BMI, fasting glucose, and insulin metrics. Predictive models will be built separately for each diet arm using penalized regression and tree-based methods to estimate both continuous weight loss and responder status, comparing omics-only, clinical-only, and integrated approaches. Candidate biomarkers will be identified through stable model features and differential expression analyses with false discovery rate control, followed by pathway enrichment to define diet-specific biological signatures. Model performance will be evaluated using repeated cross-validation, and secondary and sensitivity analyses will explore glycemic subgroups, alternative responder definitions, and robustness to adherence, while applying dimensionality-reduction and multiple-testing corrections to minimize overfitting.

5. Budget

This pilot project is entirely computational and depends on adequate hardware capacity and specialized analytical expertise to ensure rigorous, reproducible results.

High-performance computer ($3,000): Required to efficiently process large transcriptomic and proteomic datasets, run integrative bioinformatics pipelines, and perform machine-learning analyses without reliance on external servers. This ensures data security, speed, and independence for iterative analyses.
External hard drive ($200): Provides secure local backup for raw and processed omics data, supporting data integrity, redundancy, and compliance with best practices for data management.
Scientific advisor/consultant ($1,800): Supports short-term expert input on multi-omics data integration, advanced statistical modeling, and biomarker validation strategies. This guidance will strengthen analytical rigor and accelerate the identification of robust candidate signatures.

These costs are flexible and may be adjusted as pilot findings evolve. Together, this targeted investment provides the computational capacity and expert oversight necessary to complete high-quality multi-omics analyses and generate publishable, actionable pilot data.

6. References

Silverii, G., Cosentino, C., Santagiuliana, F., Rotella, F., Benvenuti, F., Mannucci, E., & Cresci, B. (2022). Effectiveness of low‐carbohydrate diets for long‐term weight loss in obese individuals: A meta‐analysis of randomized controlled trials. Diabetes, Obesity & Metabolism, 24, 1458 - 1468. https://doi.org/10.1111/dom.14709

Volek, J., Kackley, M., & Buga, A. (2024). Nutritional Considerations During Major Weight Loss Therapy: Focus on Optimal Protein and a Low-Carbohydrate Dietary Pattern. Current Nutrition Reports, 13, 422 - 443. https://doi.org/10.1007/s13668-024-00548-6

Hjorth, M., Astrup, A., Zohar, Y., Urban, L., Sayer, R., Patterson, B., Herring, S., Klein, S., Zemel, B., Foster, G., Wyatt, H., & Hill, J. (2018). Personalized nutrition: pretreatment glucose metabolism determines individual long-term weight loss responsiveness in individuals with obesity on low-carbohydrate versus low-fat diet. International Journal of Obesity, 1-8. https://doi.org/10.1038/s41366-018-0298-4

Gardner, C., Trepanowski, J., Del Gobbo, L., Hauser, M., Rigdon, J., Ioannidis, J., Desai, M., & King, A. (2018). Effect of Low-Fat vs Low-Carbohydrate Diet on 12-Month Weight Loss in Overweight Adults and the Association With Genotype Pattern or Insulin Secretion: The DIETFITS Randomized Clinical Trial. JAMA, 319, 667–679. https://doi.org/10.1001/jama.2018.0245

Ebbeling, C., Feldman, H., Klein, G., Wong, J., Bielak, L., Steltz, S., Luoto, P., Wolfe, R., Wong, W., & Ludwig, D. (2018). Effects of a low carbohydrate diet on energy expenditure during weight loss maintenance: randomized trial. The BMJ, 363. https://doi.org/10.1136/bmj.k4583

Lai, C., Parnell, L., Das, S., Gardner, C., & Ordovás, J. (2025). Differential weight‐loss responses of APOA2 genotype carriers to low‐carbohydrate and low‐fat diets: the DIETFITS trial. Obesity, 33, 1048 - 1057. https://doi.org/10.1002/oby.24288

Li, L., Li, R., Qiu, Z., Zhu, K., Li, R., Zhao, S., Che, J., Guo, T., Xu, K., Geng, T., Liao, Y., Pan, A., & Liu, G. (2025). Prediction of Weight Loss and Regain Based on Multiomic and Phenotypic Features: Results From a Calorie-Restricted Feeding Trial. Diabetes care. https://doi.org/10.2337/dc25-0728

The Right Diet Is in Your DNA: Identifying Biomarkers for Personalized Weight Loss

The Right Diet Is in Your DNA: Identifying Biomarkers for Personalized Weight Loss

Funding Progress

Funders

Peer Reviews

Related Funding

Request for Proposals - Genomics and Proteomics

Nonprofit Recipient

DOI