Table 1. Island method estimates for λ and K.
c |
Rc |
λc |
Kc |
20 | 508 087 143 | 0.2790 0.0000 (0.00%) | 0.058 |
21 | 382 046 389 | 0.2771 0.0000 (0.01%) | 0.056 |
22 | 288 047 946 | 0.2754 0.0000 (0.01%) | 0.053 |
23 | 217 666 586 | 0.2739 0.0000 (0.01%) | 0.051 |
24 | 164 854 001 | 0.2726 0.0000 (0.01%) | 0.050 |
25 | 125 090 432 | 0.2716 0.0000 (0.01%) | 0.048 |
26 | 95 080 777 | 0.2707 0.0000 (0.01%) | 0.047 |
27 | 72 367 615 | 0.2700 0.0000 (0.01%) | 0.046 |
28 | 55 135 823 | 0.2694 0.0000 (0.01%) | 0.045 |
29 | 42 040 928 | 0.2689 0.0000 (0.02%) | 0.044 |
30 | 32 087 753 | 0.2685 0.0000 (0.02%) | 0.044 |
31 | 24 502 349 | 0.2681 0.0001 (0.02%) | 0.043 |
32 | 18 721 366 | 0.2678 0.0001 (0.02%) | 0.043 |
33 | 14 312 497 | 0.2676 0.0001 (0.03%) | 0.042 |
34 | 10 945 852 | 0.2674 0.0001 (0.03%) | 0.042 |
35 | 8 372 081 | 0.2672 0.0001 (0.03%) | 0.042 |
36 | 6 407 611 | 0.2671 0.0001 (0.04%) | 0.042 |
37 | 4 904 102 | 0.2670 0.0001 (0.05%) | 0.041 |
38 | 3 755 281 | 0.2671 0.0001 (0.05%) | 0.042 |
39 | 2 874 422 | 0.2670 0.0002 (0.06%) | 0.041 |
40 | 2 201 167 | 0.2671 0.0002 (0.07%) | 0.042 |
41 | 1 684 893 | 0.2670 0.0002 (0.08%) | 0.041 |
42 | 1 289 490 | 0.2669 0.0002 (0.09%) | 0.041 |
43 | 986 932 | 0.2667 0.0003 (0.10%) | 0.041 |
44 | 756 060 | 0.2668 0.0003 (0.12%) | 0.041 |
45 | 579 087 | 0.2668 0.0004 (0.13%) | 0.041 |
46 | 443 934 | 0.2671 0.0004 (0.15%) | 0.042 |
47 | 339 913 | 0.2671 0.0005 (0.17%) | 0.042 |
48 | 260 519 | 0.2675 0.0005 (0.20%) | 0.042 |
49 | 199 117 | 0.2671 0.0006 (0.22%) | 0.042 |
50 | 152 595 | 0.2674 0.0007 (0.26%) | 0.042 |
51 | 116 705 | 0.2671 0.0008 (0.29%) | 0.042 |
52 | 89 323 | 0.2671 0.0009 (0.34%) | 0.042 |
53 | 68 605 | 0.2680 0.0010 (0.38%) | 0.044 |
54 | 52 570 | 0.2686 0.0012 (0.44%) | 0.045 |
55 | 40 242 | 0.2690 0.0013 (0.50%) | 0.046 |
56 | 30 746 | 0.2690 0.0015 (0.57%) | 0.046 |
57 | 23 481 | 0.2688 0.0018 (0.65%) | 0.046 |
58 | 17 888 | 0.2678 0.0020 (0.75%) | 0.043 |
59 | 13 662 | 0.2673 0.0023 (0.86%) | 0.042 |
60 | 10 427 | 0.2664 0.0026 (0.98%) | 0.039 |
A total of 92 441 pairs of length 7000 random sequences were generated using a set of standard amino acid frequencies (26). Island scores were generated for each pair using an extension of the Smith–Waterman algorithm (5), modified for affine gap scores (28). Substitutions were scored using the BLOSUM-62 matrix (27), and gaps of length k were assessed the score –(11 + k). Scores were recorded only for islands anchored within the central 5000 × 5000 square of each pairwise comparison. Maximum-likelihood estimates for λ and K were obtained using equations 4–6.