Skip to main content
Figure 9 | Microbial Informatics and Experimentation

Figure 9

From: Large-scale experimental studies show unexpected amino acid effects on protein expression and solubility in vivo in E. coli

Figure 9

Opposing influence of sequence parameters on protein expression/solubility vs. crystallization propensity. Predictive values for usability (i.e., having E*S > 11) were calculated using binary logistic regression against the Analysis Dataset, while predictive values for the propensity to yield a crystal structure were calculated using binary logistic regression against crystallization results obtained from a subset of the same proteins in the NESG pipeline, as previously reported [35]. Predictive values (defined in the legend for Figure 4) are compared because they normalize for the large difference in the sample sizes in the expression and crystallization datasets (9,866 vs. 679 proteins), which prevents comparison based on measures of statistical significance. Parameters are shown if significant at the indicated Bonferroni-corrected p-value thresholds in either analysis.

Back to article page

Microbial Informatics and Experimentation

ISSN: 2042-5783