Table 1.
Some existing methods for predicting protein crystallization from sequences.
Table 2.
Performance of the Init-SCM method using the p-collocated AA pairs.
Figure 1.
Heat map of the propensity scores of dipeptides obtained from the SCM method.
Table 3.
Mean performance of the SCM method using the p-collocated AA pairs.
Figure 2.
Distribution of locations of high-score dipeptides on the two typical sequences 3K9I and Q4V970.
The distribution of locations of high-score dipeptides on the two typical sequences 3K9I and Q4V970 correctly predicted as crystallizable and non-crystallizable proteins, respectively.
Table 4.
Comparisons of the proposed method SCMCRYS with existing classifiers.
Table 5.
Performances of SVM using amino acid composition (AAC) and p-collocated AA pairs.
Figure 3.
The scatter plot of correlation between solubility scores and crystallizability.
scores where R = 0.52.
Table 6.
The propensity scores of amino acids to be crystallizable and related physicochemical properties.
Table 7.
The five top-ranked physiochemical properties in the AAindex database having the highest absolute correlation with crytalizability scores of amino acids.
Figure 4.
The three-dimensional structure of Rho GDP-dissociation inhibitor.
(a) The predicted structure of a wild type Rho GDP-dissociation inhibitor and (b) The structure of a mutant Rho GDP-dissociation inhibitor (NDelta66: K135,138,141A;L196F mutant; 1fso).
Table 8.
The datasets for evaluating the predictors of protein crystallization, obtained from Mizianty and Kurgan [13].
Figure 5.
The system flowchart of the SCMCRYS method.