id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-338041-gl65i3s0	Tang, Qin	Inferring the hosts of coronavirus using dual statistical models based on nucleotide composition	2015-11-26		.txt	text/plain	4455	230	53	Both the support vector machine (SVM) model and the Mahalanobis distance (MD) discriminant model achieved high accuracies in leave-one-out cross-validation of training data consisting of 730 representative coronaviruses (99.86% and 98.08% respectively). Based on the data matrix of nucleotide composition, the MD and SVM were applied to predict hosts of coronaviruses. The data matrix with 19 factors as columns and 730 samples as rows was fitted to SVM and MD models, all predictions in leave-one-out cross-validations were listed in Supplementary Table S2 and summarized in Table 1 according to host species. Cross-host evolution research of SARS-CoV in palm civet and humans indicated that the variations in spike genes seemed to be essential for the transition of coronavirus from animal-to-human transmission to human-to-human transmission 25 . The MD correctly predicts bats as the natural hosts of the three viruses, and the SVM indicates that Rs3367 and SL-CoV-WIV1 are harmful to humans.	./cache/cord-338041-gl65i3s0.txt	./txt/cord-338041-gl65i3s0.txt