Vasina_2021_ResearchSquare__

https://www.researchsquare.com/article/rs-1027271/v1 Next-generation sequencing doubles genomic databases every 2.5 years. The accumulation of sequence data raises the need to speed up functional analysis. Herein, we present a pipeline integrating bioinformatics and microfluidics and its application for high-throughput mining of novel haloalkane dehalogenases. We employed bioinformatics to identify 2,905 putative dehalogenases and selected 45 representative enzymes, of which 24 were produced in soluble form. Droplet-based microfluidics accelerates subsequent experimental testing up to 20,000 reactions per day while achieving 1,000-fold lower protein consumption. This resulted in doubling the dehalogenation 'toolbox' characterized over three decades, yielding biocatalysts surpassing the efficiency of currently available enzymes. Combining microfluidics with modern global data analysis provided precious mechanistic information related to the high catalytic efficiency of new variants. This pipeline applied to other enzyme families can accelerate the identification of biocatalysts for industrial applications as well as the collection of high-quality data for machine learning.