Network Analysis to Identify Biological Pathways Associated with Lyme Disease

Thumbnail Image
Anand, Shreyansh
Lyme disease , Bioinformatics , Genetics
Lyme disease (LD) was the most common vector-borne pathogen in the United States from 2004-2016 and is mainly caused by the spirochete bacteria Borrelia burgdoferi sensu stricto, spreading through tick bites. It has different symptoms such as neurological issues, inflammation, and arthritis. The etiology of the disease is not well known, with a significant number of patients having long-term symptoms posttreatment with little understanding of why. Current testing is also unreliable with the most common method - the two-tier serological test - having false negative results reaching 39%. Furthermore, previous studies have focused on univariate analysis potentially missing additive gene effects. I analyzed two different datasets containing Peripheral Blood Mononuclear Cells (PBMC) based RNA-Sequencing data longitudinally collected from LD patients living in the Mid-Atlantic United States. The first dataset had 72 patients and the latter had 29, both contained data from diagnosis and post-antibiotic treatment. I used Whole Genome Correlation Network Analysis (WGCNA) along with univariate analysis and enrichment analysis to identify potential pathways of interest generating novel hypotheses for the underlying mechanisms of LD. I identified specific T-Cell-based inflammatory, arthritic and neurological pathways that were significantly elevated in LD patients and other elevated pathways that are commonly associated with symptoms found in different infectious diseases consistent with co-infections. The specific pathways and genes identified can be used as a basis for further research on the genetic mechanisms of LD. The impact of these findings can lead to the development of unique genetic biomarkers which can replace the current unreliable testing methods as well as be used for targeted therapy which can focus on the specific pathways & genes identified.
External DOI