Mutants of β-glucosidase B from Paenibacillus polymyxa
Accurately modeling the stability and catalytic function of enzymes remains a central challenge in protein design, limited largely by the scarcity of standardized experimental data available to train and benchmark computational tools. The Design to Data (D2D) program addresses this gap by engaging students worldwide to generate harmonized biophysical measurements across an expanding β-glucosidase B (BglB) variant library from Paenibacillus polymyxa. The collected papers contribute new single-point BglB variants to this shared dataset, each characterized for soluble expression, Michaelis-Menten kinetic constants (kcat, KM, and kcat/KM), and thermal stability (TM). Across the studies, variants were modeled in Foldit Standalone or related Rosetta-based tools, produced in E. coli, and assayed through colorimetric kinetic and fluorescence-based protein unfolding methods. Individual studies report a range of findings, including weak but recurring relationships between predicted total system energy (TSE) and measured stability or catalytic efficiency, the influence of active-site hydrogen bonding on expression, and mutational trends that both corroborate and contradict patterns seen in the broader dataset. Together, these contributions deepen the functional mapping of BglB and supply model-ready benchmarks for developing and evaluating the next generation of data-driven predictors of enzyme activity and stability.
N160L, N160S, N160C, N160M, and N160G | L336M, L336A, L336S, L336H, D35E, and D35W | Q22T, W123R, F155G, Y169M, W438D, and V401A | Y333F, A88E, L219Q, A408H, Y173L, E340S, and Y422F | V311D, F248N, Y166H, Y166K, and M221K | I45K, A357S, I20A, I20V, and I20E | V311D, F248N, Y166H, Y166K, and M221K | L171M, H178M, M221L, E406W, N160E, and F415M | M319C, T431I, and K337D | M319C, T431I, and K337D