id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-261111-g1qxo01i	Kowalewski, Joel	Predicting novel drugs for SARS-CoV-2 using machine learning from a >10 million chemical space	2020-08-06		.txt	text/plain	6029	349	58	There are subsequently unmet needs in COVID-19 research, including identification of compounds that target the relevant SARS-CoV-2 human proteins from (1) approved drugs, (2) FDA registered chemicals or (3) a large repository of~14 million purchasable chemicals from the ZINC 15 database [18] , which we computed additional properties for such as mammalian toxicity, vapor pressure, and logP. For 65 human protein targets that SARS-CoV-2 interacts with that had publicly available bioassay and chemical data [6] , we first generated a database of predictions based on structural similarity to chemicals that interact with the targets and then machine learning models (34) . Accordingly, we used the machine learning models to predict activities of 100,000 FDA registered chemicals (UNII database) [19] as well as the DrugBank [20] and Therapeutic Targets [21, 22] databases, which include information on drug interactions, pathways, and approval status.	./cache/cord-261111-g1qxo01i.txt	./txt/cord-261111-g1qxo01i.txt