view article

Figure 7
Confusion matrix showing the results for the sequence-based prediction of crystallizability: those sequences considered to be crystallizable (`diffraction quality crystal') and those considered noncrystallizable (`work stopped'). The results on the 2000 sequences in the independent test set show that 1477 (73.9%) can be predicted correctly, with the correctly predicted noncrystallizable and crystallizable sequences accounting for 687/2000 (34.4%) and 790/2000 (39.5%), respectively. Thus, 210/2000 (10.5%) of crystallizable and 313/2000 (15.7%) of noncrystallizable sequences were incorrectly predicted.

ISSN: 2053-230X
Follow Acta Cryst. F
Sign up for e-alerts
Follow Acta Cryst. on Twitter
Follow us on facebook
Sign up for RSS feeds