Tipton, Laura
(2017)
Quantitative Inferences from the Lung Microbiome.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Within the last decade, we have progressed from the belief that the healthy human lung is a sterile environment to attempts to study inter-kingdom interactions between microbial residents of the lungs. It has been repeatedly confirmed that the lungs contain both bacteria, predominantly from the Streptococcus, Veillonella, and Prevotella genera, and fungi, predominantly from the Cladosporium, Eurotium, Penicillium, and Aspergillus genera. The community composition as a whole undergoes shifts in every lung disease and condition that has been studied, including asthma, chronic obstructive pulmonary disorder, and cystic fibrosis. The studies that have observed these shifts have largely been descriptive, comparing the taxonomies present in healthy lungs to taxonomies in diseased lungs. Here we investigated the lung microbiome and relationships within the microbial community and between microbes and the host in a more quantitative and inferential manner. First, we introduced the lasso-penalized generalized linear mixed model (LassoGLMM) for microbiomes. LassoGLMM was applied to a short time-course study of the human oral bacterial microbiome with standard blood chemical measurements and to repeated measurements of the human lung bacterial microbiome and fungal mycobiome with local and systemic markers of inflammation. We sought to show that increased inflammation and other continuous clinical variables in human hosts are associated with distinct microbes present in the lung or oral microbiomes. Then, we examined cross-domain interactions between bacteria and fungi. Ecological interaction networks were inferred for the human lung and skin micro- and myco-biomes. Networks limited to a single domain of life were compared with those that include both bacteria and fungi to identify important components of the microbial community that would be overlooked in a single domain study. Finally, we explored the metabolism of the bacteria within the human lung using three different “-omics” datasets: taxonomic assignments from 16S rRNA gene sequences, gene families from metatranscriptomic sequences, and mass-to-charge ratio (m/z) features from metabolomics. Correlations were examined between pairs of datasets and all three datasets were integrated to identify bacteria contributing metabolic processes that may have otherwise gone unnoticed, resulting in the first complete characterization of the metabolism of the human lung bacterial microbiome.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
27 February 2017 |
| Date Type: |
Publication |
| Defense Date: |
14 December 2016 |
| Approval Date: |
27 February 2017 |
| Submission Date: |
22 December 2016 |
| Access Restriction: |
1 year -- Restrict access to University of Pittsburgh for a period of 1 year. |
| Number of Pages: |
186 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
School of Medicine > Computational Biology |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
microbiome, mycobiome, respiratory tract, penalized regression, ecological networks |
| Date Deposited: |
27 Feb 2017 20:23 |
| Last Modified: |
27 Feb 2018 06:15 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/30639 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}