|
Activity Number:
|
467
- Modeling, Design Strategies and Assessment of Biomarkers
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, July 31, 2019 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #306866
|
|
|
Title:
|
Prognostic Models from Data Integration of Clinical Characteristics and Gene Expression Data Using Bayesian Networks
|
|
Author(s):
|
Duncan Rotich* and Jeffrey A. Thompson
|
|
Companies:
|
University of Kansas Medical Center and University of Kansas Medical Center
|
|
Keywords:
|
Data Integration;
Bayesian Networks;
Survival Analysis
|
|
Abstract:
|
Recent advances in technology have generated vast amounts of heterogeneous types of data collected from patients. These data types range from clinical characteristics and patient-reported outcomes to high throughput genomic data. Currently, most studies approach these data types independently even though there might exist important relationships which can be revealed through data integration. Integrating information from multiple sources is important in capturing patient heterogeneity which consequently can help inform treatment regimens to achieve precision medicine. We propose a Bayesian Network survival model that integrates both clinical information and gene expression data for increased accuracy in prognosis. Publicly available data from The Cancer Genome Atlas(TCGA) was utilized in developing the models. Models are learned and evaluated through cross-validation and performance assessed in comparison to other standard approaches using concordance index. By incorporating expert knowledge, as well as multiple sources of data in other context, Bayesian Network algorithms can provide increased precision compared to standard approaches regarding model prediction.
|
Authors who are presenting talks have a * after their name.
