Table 1. ViroFind performance characteristics.
| Sample | Platform | Mapped JCV Reads | Total Reads | Mapped Reads/Total Reads | # Fold Enrichment |
|---|---|---|---|---|---|
| PML Brain sample 1 | ViroFind | 1,097 | 1,023,903 | 1 x 10−3 | 33 |
| Deep Sequencing | 24 | 738,541 | 3 x 10−5 | ||
| PML Brain sample 2 | ViroFind | 9,327 | 1,826,482 | 5 x 10−3 | 58 |
| Deep Sequencing | 45 | 515,119 | 8 x 10−5 | ||
| PML Brain sample 3 | ViroFind | 375,653 | 430,842 | 9 x 10−1 | 127 |
| Deep Sequencing | 2,940 | 428,463 | 7 x 10−3 | ||
| PML Brain sample 4 | ViroFind | 584 | 1,001,280 | 6 x 10−4 | 43 |
| Deep Sequencing | 7 | 520,848 | 1 x 10−5 | ||
| PML Brain sample 5 | ViroFind | 10,345 | 1,345,674 | 7 x 10−3 | 116 |
| Deep Sequencing | 56 | 899,654 | 6 x 10−5 |
The table shows the enrichment of JC virus sequences attributable to ViroFind in five brain samples from PML patients. We extracted DNA from brain tissue and made sequencing libraries that were used for a deep sequencing experiment to detect JC virus sequences. We then used the same libraries for a virus capture reaction (ViroFind). The number of unique viral reads per total number of reads increased substantially when using ViroFind and the enrichment of JC virus reads attributable to our platform was between 33–127 fold.