Table 1.
Simulations of mapping errors with MEM seeds.
| k | N | Not seeded | Seeded, not best | Relative error |
|---|---|---|---|---|
| 50 | 1 | 9.2·10−4 | 8.4·10−5 | 1.09 |
| 50 | 10 | 9.0·10−3 | 7.9·10−4 | 1.09 |
| 50 | 100 | 8.1 · 10−2 | 5.1 · 10−3 | 1.06 |
| 100 | 1 | 2.4 · 10−5 | 7.3 · 10−6 | 1.30 |
| 100 | 10 | 2.9 · 10−4 | 6.9 · 10−5 | 1.24 |
| 100 | 100 | 5.5 · 10−3 | 5.7 · 10−4 | 1.10 |
Ideal reads were randomly generated 1,000,000 times with sequencing error rate p = 0.01, divergence rate μ = 0.06, varying the read size k and number N of duplicates as indicated, using MEM seeds of minimum size γ = 19. Not seeded shows how often the read had no seed for the target, Seeded, not best shows how often the read had a seed for the target but the target was not the best candidate, and Relative error shows the factor between the probability that the read is not mapped to the true location and the probability that seeding is off-target.