An Integrated RNA and DNA Molecular Signature for Colorectal Cancer Classification
Date
2019-08
Authors
Mohammed, Mohanad
Mwambi, Henry
Omolo, Bernard
Journal Title
Journal ISSN
Volume Title
Publisher
Strathmore University
Abstract
Colorectal cancer (CRC) is the third most common cancer among women and men
in the USA. The KRAS gene is mutated in 40% of the CRC cases and hence the
RAS pathway activation has become a major focus of drug targeting efforts.
However, nearly 60% of patients with wild-type KRAS fail to respond to RAS targeted
therapies, for example the anti-epithelial growth factor receptor inhibitor
(EGFRi) combination therapies. Thus, there is a need to develop more reliable
molecular signatures to better predict mutation status. In this study, we develop a
hybrid (DNA mutation and RNA expression) signature and assess its predictive
properties for the mutation status of CRC patients. Publicly-available microarray
and RNA-Seq data from 54 matched formalin-fixed paraffin embedded (FFPE)
samples from the Affymetrix GeneChip and RNA-Seq platforms, were used to
obtain differentially expressed genes between mutant and wild-type samples. For
classification, the support-vector machines, artificial neural networks, random
forests, k-nearest neighbors and the nave Bayes algorithms were employed.
Compared to the genelist from each of the platforms, the hybrid genelist had the
highest accuracy, sensitivity, specificity and AUC for mutation status and could
therefore be useful in clinical practice, especially for colorectal cancer diagnosis
and therapeutics.
Description
Paper presented at the 5th Strathmore International Mathematics Conference (SIMC 2019), 12 - 16 August 2019, Strathmore University, Nairobi, Kenya