Table 1.

The number of viral sequences of various sizes from viral genomes discovered before January 2014, between January 2014 and May 2015, and after May 2015

Length	Training (Before 1/2014)	Validation (1/2014–5/2015)	Test (After 5/2015)	Total
150 bp	505,259	164,918	355,204	705,697
300 bp	252,630	82,458	177,416	512,504
500 bp	154,640	50,350	106,298	311,288
1000 bp	77,014	25,087	52,956	155,057
3000 bp	25,263	8,246	17,385	50,894

The three parts of the dataset partitioned by dates were used for training, validation, and testing, respectively.