Comprehensive Analysis Examining Metabolites Associated with Severity of Cystic Fibrosis 公开

Dave, Ishaan (2017)

Permanent URL: https://etd.library.emory.edu/concern/etds/s1784m577?locale=zh
Published

Abstract

Introduction: Cystic fibrosis (CF) is an inherited monogenetic disease caused by mutations in the CF transmembrane conductance regulator (CFTR) gene that affects approximately 80,000 people across the world. Altering the gene leads to improper epithelial secretions and multi-organ dysfunction. Discerning changes in airway fluid of CF infants to curb disease is of great interest. The goal of this thesis is to perform a metabolomics analysis on bronchoalveolar lavage fluid (BALF) obtained from an 11 patient cohort aged 35-38 months using untargeted mass spectrometry (MS) profiling.

Methods: Several techniques were employed to analyze data generated by MS of BALF to determine metabolites most significantly associated with severity of CF airway damage, as measured by the sensitive PRAGMA scoring tool for computed tomography images acquired concomitantly with BALF collection. MS data yielded 2,591 features in CF BALF, and Spearman and Pearson correlations for each feature with PRAGMA scores were calculated. Features with a significant Spearman correlation were run through mummichog (network analysis tool) to identify pathways in which these metabolites are involved. Penalized regression selected important metabolites among those with significant Pearson correlations to build a parsimonious predictive model for PRAGMA. Finally, a random forest algorithm was run on all the features to identify those most important in predicting PRAGMA.

Results: We identified 105 and 101 features significantly correlated with PRAGMA score using Spearman and Pearson methods, respectively. The 105 Spearman "hits" run through mummichog identified several amino acid metabolism pathways, including that of tryptophan, featuring formyl-N-acetyl-5-methoxykynurenamine (AFMK). The random forest algorithm identified 3 important features - GlcCer(d14:1(4E)/20:0(2OH)), tetrahydrocorticosterone, and AFMK. From the penalized regression utilizing the 101 Pearson hits, 11 metabolites were selected to build a predictive model for the PRAGMA score. Among them, nonate and PGF2 alpha-dihydroxypropanylamine, are particularly scientifically insightful.

Discussion: Random forest and mummichog identified AFMK - a metabolite of the tryptophan metabolism pathway implicated in the maintenance of mucosal integrity. Penalized regression identified nonate, a succinate derivative - important in the electron transfer chain of mitochondria - and a PGF2 alpha derivative, putatively linked to inflammatory signaling. These results bring novel insights into mechanisms underlying airway disease development in CF infants.

Table of Contents

Introduction ………………………………………………………………………….…...1

Problem Statement …………………………………………………………………….…2

Background/Literature Review …………………………………………………………. 3

Metabolomics Studies of Early CF Disease ……..……………………………………….4

Review of Statistical Methods for Metabolomics Data………………………….…...…….5

Methods ………………………………….……………………………………………….8

PRAGMA Data Collection………………………………………………………..8

Metabolomics Data Collection……………………………………………………9

Pathway Analysis……………………………………………………………………...…10

Statistical Data Mining……………………………………………………………………..10

Results ……………….…………………………………………………………………. 11

Discussion ………………...……………………………………………………………. 15

Limitations ...…………………………………………………………………………… 16

References ……………………………………………………………………………….17

Other references....................………………………………………...…...18

About this Master's Thesis

Rights statement
  • Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.
School
Department
Subfield / Discipline
Degree
Submission
Language
  • English
Research field
关键词
Committee Chair / Thesis Advisor
Committee Members
Partnering Agencies
最新修改

Primary PDF

Supplemental Files