gUMI-BEAR, a modular, unsupervised population barcoding method to track variants and evolution at high resolution

Abstract

Cellular lineage tracking provides the means to observe population makeup at the clonal level, allowing exploration of heterogeneity, evolutionary and developmental processes and individual clones' relative fitness. Its use, however, is limited because existing methods are highly specific, expensive, labour-intensive, and, critically, do not allow the repetition of experiments. To address these issues, we developed gUMI-BEAR (genomic Unique Molecular Identifier Barcoded Enriched Associated Regions), a modular, cost-effective method for tracking populations at high resolution. We first demonstrate the system's application and resolution by applying it to track millions of Saccharomyces cerevisiae lineages growing together across multiple generations, revealing fitness differences, lineage-specific adaptations to environmental changes and subtle dynamic shifts. Then, we demonstrate how gUMI-BEAR can be used to perform parallel screening and optimisation of virtually any number of gene variants, thus enabling unsupervised identification of individuals optimised for particular tasks. Comparison between multiple, identical libraries allowed us to reveal the interplay between stochastic and deterministic outcomes in this experiment.