Predicting Anticancer Drug Responses Using a Dual-Layer Integrated Cell Line-Drug Network Model.

Naiqian Zhang,Jun Wang,X Shirley Liu,Haiyun Wang,Xiaoqi Zheng,Yun Fang,Christina S Leslie

doi:10.1371/journal.pcbi.1004498

Abstract

The ability to predict the response of a cancer patient to a therapeutic agent is a major goal in modern oncology that should ultimately lead to personalized treatment. Existing approaches to predicting drug sensitivity rely primarily on profiling of cancer cell line panels that have been treated with different drugs and selecting genomic or functional genomic features to regress or classify the drug response. Here, we propose a dual-layer integrated cell line-drug network model, which uses both cell line similarity network (CSN) data and drug similarity network (DSN) data to predict the drug response of a given cell line using a weighted model. Using the Cancer Cell Line Encyclopedia (CCLE) and Cancer Genome Project (CGP) studies as benchmark datasets, our single-layer model with CSN or DSN and only a single parameter achieved a prediction performance comparable to the previously generated elastic net model. When using the dual-layer model integrating both CSN and DSN, our predicted response reached a 0.6 Pearson correlation coefficient with observed responses for most drugs, which is significantly better than the previous results using the elastic net model. We have also applied the dual-layer cell line-drug integrated network model to fill in the missing drug response values in the CGP dataset. Even though the dual-layer integrated cell line-drug network model does not specifically model mutation information, it correctly predicted that BRAF mutant cell lines would be more sensitive than BRAF wild-type cell lines to three MEK1/2 inhibitors tested.

Highlights

Over the past two decades, substantial improvements in high-throughput profiling technologies and systems approaches have increased expectations that personalized or precision medicine will become the paradigm of future medical science [1,2,3]
In this study, using the Cancer Cell Line Encyclopedia (CCLE) and Cancer Genome Project (CGP) studies as benchmark datasets, we explored the application of similarity information between cell lines and drugs in drug response prediction
Using CCLE and CGP studies as benchmark datasets, we evaluated the predictive power of our model and found that our dual-layer integrated cell line-drug network model is significantly better than model that use either the cell line similarity network (CSN) or the drug similarity network (DSN) layer alone, as well as the elastic net model

Summary

Introduction

Over the past two decades, substantial improvements in high-throughput profiling technologies and systems approaches have increased expectations that personalized or precision medicine will become the paradigm of future medical science [1,2,3]. In contrast to the one-size-fitsall approach that has dominated cytotoxic chemotherapy, personalized medicine exploits tumor response and vulnerability based on identified molecular traits to overcome some of the limitations associated with conventional symptoms-oriented disease diagnoses and therapies. The most important step in implementing personalized medicine will be the identification of biomarkers useful for predicting the drug response of a given patient [4,5,6]. Many studies conduct large-scale drug screenings on cultured human cell line panels to identify predictive biomarkers [7]. One of the earliest such attempts is the NCI-60 study [8,9], which included a set of 60 human cell lines and their responses to more than 100,000 chemical compounds. Drug response results for the NCI-60 dataset [10,11] revealed that different types of cancers have different drug response signatures, and that different tumors derived from the same type of cancer may have distinct molecular patterns [12]

Methods

Results

Conclusion