Combining genomic and network characteristics for extended capability in predicting synergistic drugs for cancer.

Yi Sun,Jian Fei,Zhuanbin Wu,Jun Feng,Ruling Shen,Kailin Tang,Qi Liu,Danyi Huang,Ruixin Zhu,Zhiwei Cao,Zhen Sheng,Chao Ma,Dingfeng Wu,Dandan Huang

doi:10.1038/ncomms9481

Yi Sun, Jian Fei + Show 12 more

Open Access

https://doi.org/10.1038/ncomms9481

Copy DOI

Abstract

The identification of synergistic chemotherapeutic agents from a large pool of candidates is highly challenging. Here, we present a Ranking-system of Anti-Cancer Synergy (RACS) that combines features of targeting networks and transcriptomic profiles, and validate it on three types of cancer. Using data on human β-cell lymphoma from the Dialogue for Reverse Engineering Assessments and Methods consortium we show a probability concordance of 0.78 compared with 0.61 obtained with the previous best algorithm. We confirm 63.6% of our breast cancer predictions through experiment and literature, including four strong synergistic pairs. Further in vivo screening in a zebrafish MCF7 xenograft model confirms one prediction with strong synergy and low toxicity. Validation using A549 lung cancer cells shows similar results. Thus, RACS can significantly improve drug synergy prediction and markedly reduce the experimental prescreening of existing drugs for repurposing to cancer treatment, although the molecular mechanism underlying particular interactions remains unknown.

Highlights

The identification of synergistic chemotherapeutic agents from a large pool of candidates is highly challenging
Experimental data on drug synergy were collected for 91 binary pairs derived from 14 compounds applied to the human diffuse large B-cell lymphoma (DLBCL) cell line OCI-LY3, and these data were complemented with information of the gene expression profiles of the cells perturbed with these individual compounds
To achieve flexibility for compounds without sufficient information, Ranking-system of Anti-Cancer Synergy (RACS) is performed in two steps (Fig. 1): (1) Preliminary ranking: RACS computes the synergistic potential for queried drug pairs in terms of similarities to known/labelled pairs relative to a targeted biological network; and (2) Secondary filtering: The preliminary ranking is further refined based on functional correlations between individual drugs by examining the gene expression profiles of tested cell lines

Summary

Introduction

The identification of synergistic chemotherapeutic agents from a large pool of candidates is highly challenging. Using The Cancer Genome Atlas (TCGA) data on lung adenocarcinoma and endocrine receptor (ER)-positive breast cancer, 28 to 38% of the combinations predicted using DrugComboRanker showed evidence of positive effects consistent with the published literature[9] These results suggest that there remains a large gap between the power of computational prediction and experimental validation. In the case of anticancer therapy, only a limited number of synergistic drugs have been identified, but the combinations between these drugs remain largely unknown or unexplored To manage this severe data imbalance, we establish a semisupervised learning model called Ranking-system of Anti-Cancer Synergy (RACS) to address the limited positive/labelled samples and the large set of unknown/unlabelled combinations. The framework of RACS can effectively improve drug synergy prediction for guiding experimental searching despite of the unclear synergistic mechanism

Methods

Results

Conclusion