Search
Close this search box.

BOSS-ME: Biological Outlier and Subtyping Software for ME/CFS

This project will develop a software tool to rapidly look for metabolism anomalies in an individual which might be explained by their genes. It will also look for potentially damaging genes in individuals and it will attempt to group ME/CFS patients based on their genetic and metabolic profiles.

  • Katherine Huang
  • Natalie Thomas, PhD
  • Robert Phair, PhD
  • David Ascher, PhD
  • Paul Gooley, PhD
  • Christopher Armstrong, PhD
  • Differential diagnosis paper accepted by Nature Communications Medicine.
  • Paper on metabolite and gene interactions in ME/CFS is in the submission process.
  • Review paper on applying machine learning to ME/CFS has been accepted with revisions by the Journal of Translational Medicine.
  • PhD Student Kathy Huang has submitted her thesis.
STUDY HYPOTHESIS AND DESCRIPTION

This project aims to create a tool that can quickly identify unusual metabolic patterns in individuals that might be linked to their genetic makeup. It will also search for genes that could potentially cause harm. Specifically, the tool will be used to classify patients with ME/CFS (Myalgic Encephalomyelitis/Chronic Fatigue Syndrome) based on their genetic and metabolic information.

The tool will be designed using data from over 1,000 people who reported having ME/CFS in the UK Biobank, which holds clinical data and blood samples from more than 300,000 individuals. By comparing these ME/CFS patients with others who have similar illnesses, we have been able to spot unique metabolic patterns that help identify ME/CFS in a person.

This new tool will be especially useful for doctors as it will make diagnosing ME/CFS faster and more accurate by spotting genetic and metabolic issues that might otherwise go unnoticed.

 

OBJECTIVES

Chart paper with several colorful, parallel graph lines

  1. Observe the relationship between genes and metabolism to identify outlier genetic anomalies and pathways in ME/CFS.
  2. Determine the relationship between metabolism and genes in ME/CFS as compared to other similar diseases.
  3. Develop an algorithm to rapidly extract outlier and pattern pathways of disease.
  4. Test and validate algorithm produced on a well-curated set of ME/CFS patients with metabolism and gene data.
  5. Eventually produce software to simplify this process.