Enzyme function annotation is a fundamental challenge, and numerous computational tools have been developed. However, most of these tools cannot accurately predict functional annotations, such as enzyme commission (EC) number, for less-studied proteins or those with previously uncharacterized functions or multiple activities. We present a machine learning algorithm named CLEAN (contrastive learning–enabled enzyme annotation) to assign EC numbers to enzymes with better accuracy, reliability, and sensitivity compared with the state-of-the-art tool BLASTp. The contrastive learning framework empowers CLEAN to confidently (i) annotate understudied enzymes, (ii) correct mislabeled enzymes, and (iii) identify promiscuous enzymes with two or more EC numbers—functions that we demonstrate by systematic in silico and in vitro experiments. We anticipate that this tool will be widely used for predicting the functions of uncharacterized enzymes, thereby advancing many fields, such as genomics, synthetic biology, and biocatalysis.

Proteins are the molecular machines of life with numerous applications in energy, health, and sustainability. However, engineering proteins with desired functions for practical applications remains slow, expensive, and specialist-dependent. Here we report a generally applicable platform for autonomous enzyme engineering that integrates machine learning and large language models with biofoundry automation to eliminate the need for human intervention, judgement, and domain expertise. Requiring only an input protein sequence and a quantifiable way to measure fitness, this automated platform can be applied to engineer a wide array of proteins. As a proof of concept, we engineer Arabidopsis thaliana halide methyltransferase (AtHMT) for a 90-fold improvement in substrate preference and 16-fold improvement in ethyltransferase activity, along with developing a Yersinia mollaretii phytase (YmPhytase) variant with 26-fold improvement in activity at neutral pH. This is accomplished in four rounds over 4 weeks, while requiring construction and characterization of fewer than 500 variants for each enzyme. This platform for autonomous experimentation paves the way for rapid advancements across diverse industries, from medicine and biotechnology to renewable energy and sustainable chemistry.

Expanding synthetic capabilities to routinely employ enzymes in organic solvents (OSs) is a dream for protein engineers and synthetic chemists. Despite significant advances in the field of protein engineering, general and transferable design principles to improve the OS resistance of enzymes are poorly understood. Herein, we report a combined computational and directed evolution study of Bacillus subtlis lipase A (BSLA) in three OSs (i. e., 1,4‐dioxane, dimethyl sulfoxide, 2,2,2‐trifluoroethanol) to devise a rational strategy to guide engineering OS resistant enzymes. Molecular dynamics simulations showed that OSs reduce BSLA activity and resistance in OSs by (i) stripping off essential water molecules from the BLSA surface mainly through H‐bonds binding; and (ii) penetrating the substrate binding cleft leading to inhibition and conformational change. Interestingly, integration of computational results with “BSLA‐SSM” variant library (3439 variants; all natural diversity with amino acid exchange) revealed two complementary rational design strategies: (i) surface charge engineering, and (ii) substrate binding cleft engineering. These strategies are most likely applicable to stabilize other lipases and enzymes and assist experimentalists to design organic solvent resistant enzymes with reduced time and screening effort in lab experiments.

Biocatalysis for the synthesis of fine chemicals is highly attractive but usually requires organic (co‐)solvents (OSs). However, native enzymes often have low activity and resistance in OSs and at elevated temperatures. Herein, we report a smart salt bridge design strategy for simultaneously improving OS resistance and thermostability of the model enzyme, Bacillus subtilits Lipase A (BSLA). We combined comprehensive experimental studies of 3450 BSLA variants and molecular dynamics simulations of 36 systems. Iterative recombination of four beneficial substitutions yielded superior resistant variants with up to 7.6‐fold (D64K/D144K) improved resistance toward three OSs while exhibiting significant thermostability (thermal resistance up to 137‐fold, and half‐life up to 3.3‐fold). Molecular dynamics simulations revealed that locally refined flexibility and strengthened hydration jointly govern the highly increased resistance in OSs and at 50–100 °C. The salt bridge redesign provides protein engineers with a powerful and likely general approach to design OSs‐ and/or thermal‐resistant lipases and other α/β‐hydrolases.