|
" Leveraging Bioinformatics Techniques to Support Biomarker Studies in Pediatric Inflammatory Diseases "
Tawalbeh, Shefa Mohammad
Hathout, Yetrib
| Document Type
|
:
|
Latin Dissertation
|
| Language of Document
|
:
|
English
|
| Record Number
|
:
|
1054611
|
| Doc. No
|
:
|
TL53728
|
| Main Entry
|
:
|
Tawalbeh, Shefa Mohammad
|
| Title & Author
|
:
|
Leveraging Bioinformatics Techniques to Support Biomarker Studies in Pediatric Inflammatory Diseases\ Tawalbeh, Shefa MohammadHathout, Yetrib
|
| College
|
:
|
State University of New York at Binghamton
|
| Date
|
:
|
2020
|
| Degree
|
:
|
Ph.D.
|
| student score
|
:
|
2020
|
| Note
|
:
|
149 p.
|
| Abstract
|
:
|
Blood accessible biomarkers are becoming highly attractive tools to assess disease progression and response to therapies, especially in pediatric diseases where other outcome measures remain challenging and are often subjective. Blood accessible biomarkers can provide insights about disease severity and progression and can be used as early readout to assess safety and efficacy of investigational drugs. With the availability of omics technologies, such as affinity arrays, large scale serum protein biomarkers discovery can be achieved. However, to define clinically meaningful biomarkers from large “omic” data sets, robust bioinformatics tools and rigorous statistical analyses are required. My research objective is to leverage statistical and bioinformatics approaches to define disease specific and treatment responsive biomarkers in two pediatric diseases, namely Duchenne muscular dystrophy (DMD; severity-associated biomarkers as well) and juvenile dermatomyositis (JDM). Such approaches include specialized differential expression analysis techniques, correlation-based network analysis for clustering and data reduction, linear mixed effect models, and enrichment analysis for functional grouping and pathway analyses. Tools used include limma (linear models for microarray data), WGCNA (weighted gene network-based correlation analysis), linear mixed effect models, and DAVID (database for annotation, visualization and integrated discovery). All custom pipelines were written using the R programming language for reproducibility and stability. Four aims are explored in my dissertation research. Aim 1 defines disease-specific and treatment-responsive biomarkers in DMD. Aim 2 defines disease-specific and treatment-responsive biomarkers in JDM. Aim 3 bridges serum protein biomarkers to clinical outcomes in DMD. Lastly, Aim 4 compares pharmacodynamic response to two commonly prescribed glucocorticoid drugs (deflazacort and prednisone) in DMD. The main contributions of this work include results of Aims 1 and 2 confirming previously known biomarkers but also defining novel, valuable biomarkers for both DMD and JDM. These biomarkers can be classified into major classes tied to different pathophysiological pathways including muscle injury, inflammation, and innate immune associated biomarkers. Some of these biomarkers responded to treatment while others did not. Moreover, from Aim 3, we defined an inventory of severity associated biomarkers for DMD. These biomarkers (as well as those from Aim 1) may be useful in a prognostic, predictive, pharmacodynamic, or monitoring context of use in DMD. Finally, Aim 4 helps define differences in protein expression levels in prednisone- vs deflazacort-treated DMD patients, which are linked to adverse effects. The broader impact of these research efforts is to enable discovery of clinically meaningful biomarkers that can be used to assess disease progression, response to therapy, and eventually predict later outcomes. These biomarkers could be used as surrogate outcomes to guide therapies for DMD and JDM and help with go-no-go decision making in future clinical trials.
|
| Descriptor
|
:
|
Bioinformatics
|
|
|
:
|
Biomedical engineering
|
|
|
:
|
Health sciences
|
|
|
:
|
Pathology
|
|
|
:
|
Pharmacology
|
| Added Entry
|
:
|
Hathout, Yetrib
|
| Added Entry
|
:
|
State University of New York at Binghamton
|
| |