VADR: validation and annotation of virus sequence submissions to GenBank

Alejandro A Schäffer; Eneida Hatcher; Linda Yankie; Lara Shonkwiler; J Rodney Brister; Ilene Karsch-Mizrachi; Eric P Nawrocki

We manually examined the remaining 21 sequences and determined that 3 of them have undeniable serious errors and so should be classified as 'negatives'. FV536857.1, which matches best to a non-norovirus RefSeq in VADR's model library (NC 006875) and appears to be reverse complemented, is failed by all three programs and so is counted for all three as a true negative (TN). FJ446720.1, which has an early in-frame stop codon 690 nt 5' of the expected stop site, is appropriately failed by VADR and VIGOR (TN) but passed by VAPiD (FP). KY905331.1 lacks a valid start codon in the nonstructural polyprotein CDS and is failed by VADR and VAPiD (TN) but passed by VIGOR (FP). We cautiously defined the remaining 18 sequences, all of which are failed by VADR but passed by VIGOR and VAPiD, as positives because each sequence is arguably valid but is just too different from the matching RefSeq model to pass VADR. These 18 sequences are TPs for VAPiD and VIGOR and FNs for VADR.