The assembly of large recombinant DNA encoding a whole biochemical pathway or genome represents a significant challenge. Here, we report a new method, DNA assembler, which allows the assembly of an entire biochemical pathway in a single step via in vivo homologous recombination in Saccharomyces cerevisiae . We show that DNA assembler can rapidly assemble a functional d -xylose utilization pathway (∼9 kb DNA consisting of three genes), a functional zeaxanthin biosynthesis pathway (∼11 kb DNA consisting of five genes) and a functional combined d -xylose utilization and zeaxanthin biosynthesis pathway (∼19 kb consisting of eight genes) with high efficiencies (70–100%) either on a plasmid or on a yeast chromosome. As this new method only requires simple DNA preparation and one-step yeast transformation, it represents a powerful tool in the construction of biochemical pathways for synthetic biology, metabolic engineering and functional genomics studies.

Actinobacteria, particularly those of genus Streptomyces, remain invaluable hosts for the discovery and engineering of natural products and their cognate biosynthetic pathways. However, genetic manipulation of these bacteria is often labor and time intensive. Here, we present an engineered CRISPR/Cas system for rapid multiplex genome editing of Streptomyces strains, demonstrating targeted chromosomal deletions in three different Streptomyces species and of various sizes (ranging from 20 bp to 30 kb) with efficiency ranging from 70 to 100%. The designed pCRISPomyces plasmids are amenable to assembly of spacers and editing templates via Golden Gate assembly and isothermal assembly (or traditional digestion/ligation), respectively, allowing rapid plasmid construction to target any genomic locus of interest. As such, the pCRISPomyces system represents a powerful new tool for genome editing in Streptomyces.

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.