We acknowledge several limitations in the current study. Although 2019-nCoV/SARS-CoV-2 shared high nucleotide sequence identity with other HCoVs (Fig. 2), our predictions are not 2019-nCoV/SARS-CoV-2 specific by lack of the known host proteins on 2019-nCoV/SARS-CoV-2. We used a low binding affinity value of 10 μM as a threshold to define a physical drug–target interaction. However, a stronger binding affinity threshold (e.g., 1 μM) may be a more suitable cut-off in drug discovery, although it will generate a smaller drug–target network. Although sizeable efforts were made for assembling large scale, experimentally reported drug–target networks from publicly available databases, the network data may be incomplete and some drug–target interactions may be functional associations, instead of physical bindings. For example, Silvestrol, a natural product from the flavagline, was found to have antiviral activity against Ebola75 and Coronaviruses76. After adding its target, an RNA helicase enzyme EIF4A76, silvestrol was predicted to be significantly associated with HCoVs (Z = –1.24, P = 0.041) by network proximity analysis. To increase coverage of drug–target networks, we may use computational approaches to systematically predict the drug-target interactions further25,26. In addition, the collected virus–host interactions are far from completeness and the quality can be influenced by multiple factors, including different experimental assays and human cell line models. We may computationally predict a new virus–host interactome for 2019-nCoV/SARS-CoV-2 using sequence-based and structure-based approaches77. Drug targets representing nodes within cellular networks are often intrinsically coupled with both therapeutic and adverse profiles78, as drugs can inhibit or activate protein functions (including antagonists vs. agonists). The current systems pharmacology model cannot separate therapeutic (antiviral) effects from those predictions due to lack of detailed pharmacological effects of drug targets and unknown functional consequences of virus–host interactions. Comprehensive identification of the virus–host interactome for 2019-nCoV/SARS-CoV-2, with specific biological effects using functional genomics assays79,80, will significantly improve the accuracy of the proposed network-based methodologies further.