Page 55 - TD-4-1

Tumor Discovery

ORIGINAL RESEARCH ARTICLE
Systemic drug repurposing for pancreatic

cancer based on genetic and epigenetic
network analysis using a systems biology

approach and deep neural learning of
drug-target interactions

Yi-Hsin Tsai and Bor-Sen Chen*
Laboratory of Automatic Control, Signal Processing and Systems Biology, Department of Electrical
Engineering, Institute of Electronic Engineering, National Tsing Hua University, Hsinchu, Taiwan,
China

Abstract

Pancreatic cancer is a malignant tumor associated with a high mortality rate.
This research presents a systems biology approach to explore the mechanisms of
pancreatic ductal adenocarcinoma (PDAC), aiming to identify significant biomarkers
that can serve as drug targets. We propose a systematic drug repurposing strategy
*Corresponding author: that incorporates a deep neural network (DNN)-based drug-target interaction (DTI)
Bor-Sen Chen
(bschen@ee.nthu.edu.tw) model along with drug design specifications to develop a potential multi-molecule
drug for PDAC treatment. We first established candidate protein-protein interaction
Citation: Tsai Y, Chen B. Systemic
drug repurposing for pancreatic networks and gene regulatory networks using big data mining techniques. Real
cancer based on genetic and PDAC and non-PDAC genome-wide genetic and epigenetic networks (GWGENs) were
epigenetic network analysis using systematically identified using their corresponding microarray data through system
a systems biology approach and
deep neural learning of drug- identification and system order detection methods. The top 6,000 core GWGENs of
target interactions. Tumor Discov. PDAC and non-PDAC were extracted using the Principal Network Projection method.
2025;4(1):47-67. Subsequently, we annotated the core GWGENs using the Kyoto Encyclopedia of
doi: 10.36922/td.4709
Genes and Genomes pathways to construct their respective core signaling pathways.
Received: August 30, 2024 By comparing upstream microenvironmental factors, core signaling pathways,
Revised: October 8, 2024 and downstream aberrant cellular functions between PDAC and non-PDAC, we
investigated the carcinogenic mechanisms of PDAC. Notably, c-MYC, forkhead box
Accepted: October 24, 2024
O3, and tumor suppressor p53 were identified as significant biomarkers for potential
Published online: November 20, drug targets. Furthermore, the DNN-based DTI model predicted the interaction
2024
probabilities between candidate molecular drugs and these biomarkers. Based
Copyright: © 2024 Author(s). on drug design specifications such as regulatory ability, sensitivity, and toxicity,
This is an Open-Access article suitable multi-molecular potential drugs were selected. Ultimately, gemcitabine and
distributed under the terms of the
Creative Commons Attribution MK-2206 were identified as a promising multi-molecular drug combination for PDAC
License, permitting distribution, treatment.
and reproduction in any medium,
provided the original work is
properly cited. Keywords: Pancreatic cancer mechanisms; Systems biology; Big data mining; Genome-
Publisher’s Note: AccScience wide genetic and epigenetic networks; Kyoto Encyclopedia of Genes and Genomes
Publishing remains neutral with pathways; Deep neural network-based drug-target interaction model; Drug design
regard to jurisdictional claims in
published maps and institutional specifications; Principal network projection
affiliations.

Volume 4 Issue 1 (2025) 47 doi: 10.36922/td.4709