|
Activity Number:
|
653
- Machine Learning and Other Statistical Methods in Clinical Trials
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Thursday, August 1, 2019 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #306757
|
Presentation
|
|
Title:
|
Application of CART Regression in Early Discovery Efforts to Better Understand Proinsulin as Possible Therapeutic Target
|
|
Author(s):
|
Santosh Sutradhar* and Geoffrey Walford and Tami Crumley and Anita Lee and Jennifer Abrams
|
|
Companies:
|
Merck & Co., Inc. and Merck & Co., Inc. and Merck & Co., Inc. and Merck & Co., Inc. and Merck & Co., Inc.
|
|
Keywords:
|
unsupervised cluster analysis;
CART;
machine learning technique
|
|
Abstract:
|
In early discovery, scientists focus on initial research generating data supporting a therapeutic target against a disease or clinical condition with unmet medical need. Statisticians play a big role helping the scientist in understanding the relationship of a particular target with the disease of interest. In early discovery efforts, we are interested in learning whether changes in proinsulin can be used for hypothesis generation toward new therapeutics for diabetes. Towards this goal, the first step is to understand the relationship of proinsulin level in patients with different disease stages. When disease stages is known, supervised cluster analysis can be used to understand the relationship. However, information on actual disease stage may not be always available. In this situation, unsupervised cluster analysis of proinsulin levels with patient characteristics may provide a surrogate for disease staging. I will discuss how an application of unsupervised machine learning technique, such as, CART regression can be useful in understanding the association of proinsulin levels with possible disease stages to evaluate proinsulin as possible therapeutic target.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2019 program
|
