Cluster learning-assisted directed evolution

Qiu, Yuchi; Hu, Jian; Wei, Guo-Wei

doi:10.1038/s43588-021-00168-y

Cited by 46 publications

(51 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several adaptive experimental design approaches, such as MLDE (17), cluster learning-assisted directed evolution (CLADE) (61), and ODBO (62), have utilized ML models for sampling-efficient protein engineering. There are two key differences between these approaches and our method.…”

Section: Discussionmentioning

confidence: 99%

“…First, BO-EVO does not require prior knowledge of the structure or homolog sequences of a target protein, which may be unavailable or scarce. On the contrary, information derived from protein structures (e.g., ΔΔG) and homolog sequences were utilized by MLDE (17) and CLADE (61), respectively, as zero-shot predictors to exclude possible zero- or low-fitness mutants from experimental sets. Second, BO-EVO does not need to evaluate the whole design space in silico by restricting sequence design via evolutionary algorithms.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Protein engineering via Bayesian optimization-guided evolutionary algorithm and robotic experiments

Chen

et al. 2022

Preprint

View full text Add to dashboard Cite

Protein engineering aims to find top functional sequences in a vast design space. For such an expensive black-box function optimization problem, Bayesian optimization is a principled sample-efficient approach, which is guided by a surrogate model of the objective function. Unfortunately, Bayesian optimization is computationally intractable with the vast search space. Even worse, it proposes sequences sequentially, making it incompatible with batched wet-lab measurement. Here, we report a scalable and batched method, Bayesian Optimization-guided EVOlutionary (BO-EVO) algorithm, to guide multiple rounds of robotic experiments to explore protein fitness landscapes of combinatorial mutagenesis libraries. We first examined various design specifications based on an empirical landscape of protein G domain B1. Then, BO-EVO was successfully generalized to another empirical landscape of an Escherichia coli kinase PhoQ, as well as simulated NK landscapes with up to moderate epistasis. This approach was then applied to guide robotic library creation and screening to engineer enzyme specificity of RhlA, a key biosynthetic enzyme for rhamnolipid biosurfactants. A 4.8-fold improvement in producing a target rhamnolipid congener was achieved after examining less than 1% of all possible mutants after 4 iterations. Overall, BO-EVO proves to be an efficient and general approach to guide combinatorial protein engineering without prior knowledge.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Protein engineering via Bayesian optimization-guided evolutionary algorithm and robotic experiments

Chen

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…By normalizing the fitness to its global maximum, 92% of sequences have fitness lower than 0.01 and 99.3% sequences have fitness lower than 0.3 for GB1. Similarly, there are 92% of sequences having fitness lower than 0.01 and 99.96% of sequences having fitness lower than 0.3 for PhoQ (Supporting Information Figure S1) …”

Section: Methodsmentioning

confidence: 96%

“…In this work, we use two combinatorial libraries, GB1 and PhoQ, that have almost complete coverage for mutations at four mutational sites. GB1 is a very popular benchmark library, while the PhoQ library has also been used in early MLDE studies. , PhoQ is considered as an alternative data set. For both data sets, their fitness values were normalized into the range [0, 1] when being applied to CLADE.…”

Section: Methodsmentioning

confidence: 99%

“…To avoid the inefficient random sampling in the huge mutational space containing a large portion of non-functional sequences, ftMLDE uses the evolutionary density models to rank sequences and confine the sampling within an informative subspace . Additionally, cluster learning-assisted DE (CLADE) uses unsupervised hierarchical clustering to guide the sampling within more informative subspace with accumulated knowledge of the fitness landscape …”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

CLADE 2.0: Evolution-Driven Cluster Learning-Assisted Directed Evolution

Qiu

Wei

2022

J. Chem. Inf. Model.

Self Cite

View full text Add to dashboard Cite

Directed evolution, a revolutionary biotechnology in protein engineering, optimizes protein fitness by searching an astronomical mutational space via expensive experiments. The cluster learning-assisted directed evolution (CLADE) efficiently explores the mutational space via a combination of unsupervised hierarchical clustering and supervised learning. However, the initial-stage sampling in CLADE treats all clusters equally despite many clusters containing a large portion of non-functional mutations. Recent statistical and deep learning tools enable evolutionary density modeling to access protein fitness in an unsupervised manner. In this work, we construct an ensemble of multiple evolutionary scores to guide the initial sampling in CLADE. The resulting evolutionary score-enhanced CLADE, called CLADE 2.0, efficiently selects a training set within a small informative space using the evolution-driven clustering sampling. CLADE 2.0 is validated by using two benchmark libraries both having 160,000 sequences from four-site mutational combinations. Extensive computational experiments and comparisons with existing cutting-edge methods indicate that CLADE 2.0 is a new state-of-art tool for machine learning-assisted directed evolution.

show abstract

Gene Mutagenesis Methods in Directed Evolution and Rational Enzyme Design

2023

Enzyme Engineering

View full text Add to dashboard Cite

Cluster learning-assisted directed evolution

Cited by 46 publications

References 51 publications

Protein engineering via Bayesian optimization-guided evolutionary algorithm and robotic experiments

Protein engineering via Bayesian optimization-guided evolutionary algorithm and robotic experiments

CLADE 2.0: Evolution-Driven Cluster Learning-Assisted Directed Evolution

Gene Mutagenesis Methods in Directed Evolution and Rational Enzyme Design

Contact Info

Product

Resources

About