Skip to main content
. Author manuscript; available in PMC: 2023 Jan 17.
Published in final edited form as: Nat Biotechnol. 2021 Feb 23;39(6):747–753. doi: 10.1038/s41587-021-00839-1

Table 1.

Overview of viral targets used for modeling and summary statistics of model identified active compounds.

Model Data Source Data Type Active
Enrichment
(fold)* 		Potency
Range** 		Structure
Diversity***
SARS-CoV-2 In house/Literature HTS/Literature 2.8-3.3 0.5 μM - 32 μM 0.07-0.23
ZIKV NS1 a In house qHTS 2.7-27.5 1 nM - 63 μM 0.04-0.26
EBOV Literature HTS 6.5-6.8 0.4 μM - 25 μM 0.07-0.21
a

The NS1 protein is a nonstructural protein that is not present in the virus itself but only appears in the host cells when virus starts replication. The NS1 assay detects compounds that block NS1 protein production, which may inhibit virus entry to cells, and virus RNA and virus protein productions in cells.

*

Fold = hit rate of model predicted actives/hit rate of assay used for model training

**

Potency range of experimentally confirmed model predicted actives

***

Tanimoto similarity was calculated between each model predicted active and active compounds in the training set. The values shown here are the range of the average Tanimoto scores for the model predicted actives.