PharmaDx Clinical Metabolomics Analysis Report

summary(cars)

##      speed           dist       
##  Min.   : 4.0   Min.   :  2.00  
##  1st Qu.:12.0   1st Qu.: 26.00  
##  Median :15.0   Median : 36.00  
##  Mean   :15.4   Mean   : 42.98  
##  3rd Qu.:19.0   3rd Qu.: 56.00  
##  Max.   :25.0   Max.   :120.00

Executive Summary

This report presents the results of an untargeted metabolomics analysis comparing three clinical groups: Healthy controls (n=20), PreDiabetic patients (n=20), and Type 2 Diabetes (T2D) patients (n=20). Fifty metabolites were profiled across 60 samples.

Key findings:

• 22 metabolites were significantly altered between Healthy and PreDiabetic groups (fully reliable comparison)
• 43 metabolites differed between PreDiabetic and T2D groups
• Beta_hydroxybutyrate, Lactate and Acetoacetate were the top-ranked biomarker candidates by both statistical significance and VIP score
• A complete batch confounding issue was identified — see Section 3

⚠️ Important limitation: All Healthy samples were analysed in Batch1 and all T2D samples in Batch2. Healthy vs T2D comparisons are batch-confounded and should be interpreted with caution. The Healthy vs PreDiabetic comparison is not affected.

Conclusions

Metabolomics profiling of 60 clinical samples identified 48 significantly altered metabolites across the disease progression from Healthy to T2D after Benjamini-Hochberg correction.

The Healthy vs PreDiabetic comparison — our most reliable finding — identified 22 significantly elevated metabolites, led by Beta_hydroxybutyrate (p = 8.96e-07), indicating that ketone body dysregulation is an early event in diabetes development, preceding clinical T2D diagnosis.

PCA and PLS-DA confirmed complete metabolic separation between all three clinical groups, supporting the biological validity of the observed metabolite changes.

Recommendation: Given the batch confounding identified between Healthy and T2D groups, we recommend the client reanalyse a subset of Healthy samples alongside T2D samples in a new analytical batch to confirm the Healthy vs T2D findings with a batch-balanced design.

---

Report prepared by Augustine Onuh, PhD | MetaboInsight Analytics Generated: 2026-08-09 | R version 4.6.0