Table 1.
For each fragment length (training dataset), the generated distribution of BackboneScores is computed and summarized via its mean, median, minimum, and maximum values.
FL | Vanilla GAN | |||
---|---|---|---|---|
Mean | Med | Min | Max | |
128 | 125.31 | 126 | 94 | 127 |
64 | 62.82 | 63 | 48 | 63 |
16 | 12.75 | 13 | 9 | 15 |
9 | 4.02 | 4 | 0 | 6 |
6 | 2.93 | 3 | 0 | 4 |
FL | Vanilla GAN + SpecNorm | |||
Mean | Med | Min | Max | |
128 | 125.04 | 126 | 101 | 127 |
64 | 62.83 | 63 | 18 | 63 |
16 | 12.62 | 13 | 11 | 14 |
9 | 5.99 | 6 | 3 | 6 |
6 | 0.01 | 0 | 0 | 2 |
FL | Vanilla GAN + VBN | |||
Mean | Med | Min | Max | |
128 | 127.00 | 127.0 | 127 | 127 |
64 | 49.08 | 61 | 0 | 63 |
16 | 15.00 | 15 | 13 | 15 |
9 | 7.94 | 8 | 4 | 8 |
6 | 4.98 | 5 | 3 | 5 |
FL | WGAN | |||
Mean | Med | Min | Max | |
128 | 51.94 | 49 | 0 | 127 |
64 | 25.56 | 24 | 0 | 63 |
16 | 0.005 | 0 | 0 | 10 |
9 | 5.52 | 6 | 0 | 8 |
6 | 0.47 | 0 | 0 | 5 |