Table 2.
Comparison of Targeted Sequencing Data Analysis by GENE-IS Targeted Sequencing Mode and Virus-Clip by Using an In Silico Dataset of 5,600 Reads and a Control Sample, Lentiviral Vector-Transduced HeLa Cells, with Three Pre-determined Integration Sites
| Tools | GENE-IS | Virus-Clip |
|---|---|---|
| 5,600-Read In Silico Dataset | ||
| Sensitivity | 1 | –a |
| Specificity | 1 | |
| Precision | 1 | |
| Accuracy | 1 | |
| Real Control HeLa Sample, ∼37.5 Million Reads | ||
| With sequence count threshold per integration site >1 | ||
| Sensitivity | 1 | 1 |
| Specificity | 1 | 0.063 |
| Precision | 1 | 0.038 |
| Accuracy | 1 | 0.097 |
| Without sequence count threshold per integration site | ||
| Sensitivity | 1 | 1 |
| Specificity | 0.28 | 0.059 |
| Precision | 0.19 | 0.036 |
| Accuracy | 0.38 | 0.091 |
Statistical measures were estimated without using any sequence count threshold per integration site and by using a threshold greater than 1, i.e., in this case only those integration sites were counted that were supported by at least two independent reads.
a
Analysis didn’t report each individual genomic IS.