Table 1

CA algorithm assignment success
m s 1-domain 2-domain 3-domain 4-domain Overall1
15 5 70% (0.49) 59% (0.34) 47% (0.28) 35% (0.21) 67%
22 5 75% (0.52) 58% (0.35) 45% (0.29) 32% (0.21) 69%
30 5 88% (0.52) 46% (0.37) 32% (0.31) 22% (0.21) 76%
22 3 63% (0.50) 49% (0.24) 42% (0.18) 32% (0.13) 58%
22 5 75% (0.52) 58% (0.35) 45% (0.29) 32% (0.21) 69%
22 8 87% (0.44) 44% (0.33) 26% (0.27) 14% (0.16) 75%
Random (see text) 76% (≈0) 15% (≈0) 2% (≈0) 0% (≈0) 60%

Only buried α-carbons were clustered, and adjacency constraint was enforced in all cases. Success given on the ASTRAL30 data set as a function of m and s. Matthews correlation coefficient is given in parentheses. 1Total, regardless of number of domains.

Feldman

Feldman BMC Bioinformatics 2012 13:286   doi:10.1186/1471-2105-13-286

Open Data