Table 5.
Stability of network inference against noise
| Real data | 1 ppm | 2 ppm | 5 ppm | 10 ppm | 20 ppm | 100 | 1000 | 10,000 | Random | |
|---|---|---|---|---|---|---|---|---|---|---|
| Clusters | 2472 | 2254 | 2040 | 1660 | 1365 | 1113 | 470 | 53 | 35 | 19 |
| Explained distances | 25,370 | 22,988 | 21,094 | 16,966 | 13,213 | 10,003 | 3261 | 328 | 213 | 115 |
| % Excess over random | 100 | 90.5 | 83.1 | 66.7 | 51.9 | 39.2 | 12.5 | 0.8 | 0.4 | 0 |
Uniformly distributed random noise of the indicated size was added to all observed masses and the network reconstructed as described in the text.