RAman SPectrometry Based biomarkER discoveRY for Myalgic Encephalomyelitis (RASPBERRY-ME)

  • The Raman measurements for the ME/CFS project are completely done. For the entire dataset, we now have five separate subgroups based on the microRNA classification:
    • Group 1: (T0 = 10, T90 = 10) patients
    • Group 2A: (T0 = 30, T90 = 30) patients
    • Group 2B: (T0 = 30, T90 = 30) patients
    • Group 3: (T0 = 30, T90 = 30) patients
    • Group 4: (T0 = 15, T90 = 15) patients
    • Healthy controls (HC): (T0 = 23, T90 = 23) participants
  • Our preliminary models show our capacity to predict with a relatively high accuracy which plasma samples are linked to M/CFS vs HC group at T90 (after the stress-test) with a ROC curve showing an area under the curve (AUC) of 0.86. Raman profiles of plasma can distinguish between healthy individuals and ME/CFS patients with high accuracy (86%).
  • The data have been sent for machine learning (ML) modeling and optimization including the development of models for each subgroup and predicting disease severity.

STUDY HYPOTHESIS AND DESCRIPTION

Woman researcher working with a computer in a laboratory. Raman spectroscopy is a non-destructive, rapid, and low-cost technique allows the study of the molecular composition of biological fluids like blood, or inside a cell when combined with confocal microscopy. This innovative approach could lead to the development of diagnostic tools to better stratify ME patients and find the underlying causes of different symptoms like post-exertional malaise as well as clinical tools to validate the therapeutic potential of pharmacological treatments to treat, stop or mitigate ME through precision medicine.

We hypothesize that our approach will allow the identification of a biomolecular signature of ME both at baseline and in response to the application of a post-exertional stress challenge. We expect to stratify patients by differentiating severe cases from mild forms of ME. Results from this study will be further combined to ongoing proteomic and metabolomic profiling approaches to better understand the pathophysiology of ME.

Objective

  1. Characterize the biomolecular signature of ME patients and healthy age-matched controls using label-free Raman spectroscopy in plasma samples acquired at baseline and after PEM induction
  2. Determine cellular metabolite alterations between ME patients and age-matched controls using Raman spectroscopy combined with confocal microscopy with peripheral blood mononuclear cells (PBMCs) acquired at baseline and after PEM induction
  3. Develop molecular feature detection and machine-learning models capable to predict PEM response (baseline vs. post stress-test) and the disease (ME vs. control).