Deep learning-based prediction of enzyme optimal pH and design of point mutations to improve acid resistance

Abstract An accurate deep learning predictor of enzyme optimal pH is essential to quantitatively describe how pH influences the enzyme catalytic activity. Seq2pHopt-2.0, developed in this study, outperformed existing predictors of enzyme optimal pH (RMSE=0.833 and R2=0.479), and could provide good interpretability with informative residue attention weights. The accurate classification of acidic and alkaline enzymes showcased the potential of Seq2pHopt-2.0 as a useful enzyme mining tool for identifying candidate enzymes with specific pH preferences. Furthermore, a single point mutation designed with the guidance of Seq2pHopt-2.0 successfully enhanced the activity of Pyrococcus horikoshii diacetylchitobiose deacetylase at low pH (pH=4.5/5.5) by approximately 7%, suggesting that Seq2pHopt-2.0 is a promising in-silico enzyme design tool for pH-dependent enzyme activities. Graphical abstract