Evaluating molecular modeling tools for thermal stability using an independently generated dataset

Abstract

Engineering proteins to enhance thermal stability is a widely utilized approach for creating industrially relevant biocatalysts. Computational tools that guide these engineering efforts remain an active area of research with new data sets and develop algorithms. To aid in these efforts, we are reporting an expansion of our previously published data set of mutants for a -glucosidase to include both measures of TM and G, to complement the previously reported measures of T50 and kinetic constants (kcat and KM). For a set of 51 mutants, we found that T50 and TM are moderately correlated with a Pearson correlation coefficient (PCC) of 0.58, indicated the two methods capture different physical features. The performance of predicted stability using five computational tools are also evaluated on the 51 mutants dataset, none of which are found to be strong predictors of the observed changes in T50, TM, or G. Furthermore, the ability of the five algorithms to predict the production of isolatable soluble protein is examined, which revealed that Rosetta ΔΔG, ELASPIC, and DeepDDG are capable of predicting if a mutant could be produced and isolated as a soluble protein.These results further highlight the need for new algorithms for predicting modest, yet important, changes in thermal stability as well as a new utility for current algorithms for prescreening designs for the production of soluble mutants. ASSOCIATED CONTENT Supporting Information (SI)SI 1-1. SDS-PAGE images for 51 BglB mutants and WT. (PDF) SI 1-2. A distribution analysis of temperatures observed for TM and T50. SI 1-3. PPC graph between ΔTM and ΔΔG of BglB mutants. (PDF) SI 1-4. Evaluation of five computational methods on protein expression. (PDF) SI 2. Images of TM fluorescence graphs, derivative graphs, and Van't Hoff plot for 51 mutants and WT. (Zip) SI 3. Rosetta ΔΔG and FoldX PSSM correlations graphs with ΔTM and experimental ΔΔG. Excel files of all the parameters from data acquisition. Excel files of the total system energy for DeepDDG, ELASPIC, and PoPMuSiC (Zip)SI 4. Jupyter notebook for all thermal stability data acquisitions with all TM raw data files. (ipynb) SI 5. Example files for Rosetta_ddg_monomer run. (Zip)

Design to Data for mutants of β-glucosidase B fromPaenibacillus polymyxa: Q22T, W123R, F155G, Y169M, W438D, V401A

“…The mean distribution of TM, the temperature at which half the protein molecules denatures, of wildtype BglB is determined to be 44.54 °C, which is in agreement with previously observed TM. 9 The TM of the Q22T mutation did not differ significantly from that of the wildtype BglB (two sample T test, p = 0.70). There was a significant difference between the mean TM of W123R and the wildtype.…”

Design to Data for mutants of β-glucosidase B fromPaenibacillus polymyxa:L171M, H178M, M221L, E406W, N160E, F415M

“…Pearson correlation coefficient (PCC) analysis was used to examine the relationship between the change in TSE to T M and k cat /K M . 14…”

PRO-LDM: Protein Sequence Generation with a Conditional Latent Diffusion Model

“…The field of protein engineering has undergone significant developments in multiple fields including site-directed mutagenesis, rational design(1), directed-evolution (2)(3)(4), semi-rational design (5)(6)(7), and computational design (8)(9)(10). Compared with traditional protein design methods, computational protein design can provide clear modification plans, greatly reduce the workload required to establish and screen mutant libraries, and made remarkable progresses in protein de novo design (11)(12)(13), enzyme substrate selectivity (14)(15)(16), thermal stability design (17)(18)(19), etc.…”