Skip to main content
. 2004 Jul 21;112(16):1614–1621. doi: 10.1289/txg.7105

Table 2.

Results from 100,000 Monte Carlo simulations of four hypothetical four-gene networks (A, B, C, D)a describing the ability of the TAO-Gen algorithm to specify the correct network.

Frequency (%) of resulting optimal network structure
Rank (%) of the posterior likelihood for the true network over all possible 543 acyclic networks
Sample size True model graphic file with name ehp0112-001614f5.jpg graphic file with name ehp0112-001614f6.jpg graphic file with name ehp0112-001614f7.jpg graphic file with name ehp0112-001614f8.jpg graphic file with name ehp0112-001614f9.jpg graphic file with name ehp0112-001614f10.jpg graphic file with name ehp0112-001614f11.jpg 1 2 3 4–10
100 arrays A 922 (92) 0 (0) 0 (0) 0 (0) 0 (0) 68 (7) 0 (0) 922 (92) 52 (5) 10 (1) 16 (2)
1,000 sims B 977 (98) 0 (0) 0 (0) 0 (0) 0 (0) 6 (1) 0 (0) 977 (98) 17 (2) 4 (0.4) 2 (0.2)
C 929 (93) 0 (0) 0 (0) 0 (0) 0 (0) 71 (7) 0 (0) 929 (93) 50 (5) 8 (1) 13 (1)
D 980 (98) 0 (0) 0 (0) 0 (0) 0 (0) 6 (1) 0 (0) 980 (98) 13 (1) 5 (0.5) 2 (0.2)
50 arrays A 1,716 (86) 4 (0.2) 3 (0.2) 6 (0.3) 4 (0.2) 165 (8) 0 (0) 1,716 (87) 157 (8) 34 (2) 70 (4)
2,000 sims B 1,841 (92) 8 (0.4) 0 (0) 4 (0.2) 8 (0.4) 41 (2) 0 (0) 1,841 (92) 82 (4) 20 (1) 55 (3)
C 1,745 (87) 6 (0.3) 4 (0.2) 3 (0.2) 6 (0.3) 175 (9) 0 (0) 1,745 (88) 128 (6) 41 (2) 62 (3)
D 1,860 (93) 4 (0.2) 0 (0) 2 (0.1) 0 (0) 46 (2) 0 (0) 1,860 (93) 68 (3) 30 (2) 42 (2)
25 arrays A 2,920 (73) 76 (2) 72 (2) 56 (1) 77 (2) 328 (8) 3 (0.1) 2,920 (73) 423 (10) 112 (3) 387 (10)
4,000 sims B 3,179 (80) 92 (2) 55 (1) 48 (1) 47 (1) 192 (5) 8 (0.2) 3,179 (79) 348 (9) 133 (3) 249 (6)
C 2,891 (72) 60 (1) 100 (2) 56 (1) 76 (2) 296 (7) 4 (0.1) 2,891 (72) 404 (10) 114 (3) 444 (11)
D 3,086 (77) 76 (2) 96 (2) 48 (1) 48 (1) 164 (4) 8 (0.2) 3,086 (77) 328 (8) 149 (4) 365 (9)
10 arrays A 3,198 (32) 909 (9) 741 (7) 230 (2) 149 (2) 328 (3) 497 (5) 3,198 (32) 1,027 (10) 781 (8) 2,389 (24)
10,000 sims B 3,768 (38) 1,002 (10) 1,051 (10) 220 (2) 309 (3) 378 (4) 567 (6) 3,768 (38) 966 (10) 821 (8) 2,519 (25)
C 3,177 (32) 892 (9) 691 (7) 230 (2) 151 (2) 398 (4) 457 (5) 3,177 (32) 1,232 (12) 769 (8) 2,347 (23)
D 3,768 (38) 1,052 (10) 1,031 (10) 280 (3) 259 (3) 538 (5) 477 (5) 3,768 (38) 1,146 (11) 871 (9) 2,371 (24)
a

(A) β14 = 2.0, β13 = 0.8, β23 = 0.8, β34 = −1.3, σ1 = σ2 = σ3 = σ4 = 1.0

(B) β14 = 2.0, β13 = 0.8, β23 = 0.8, β34 = −5.0, σ1 = σ2 = σ3 = σ4 = 1.0

(C) β14 = 2.0, β13 = 0.8, β23 = 0.8, β34 = −1.3, σ1 = σ2 = σ3 = σ4 = 1/3

(D) β14 = 2.0, β13 = 0.8, β23 = 0.8, β34 = −5.0, σ1 = σ2 = σ3 = σ4 = 1/3

a