A framework for evaluating edited cell libraries created by massively parallel genome engineering

Abstract

Genome engineering methodologies are transforming biological research and discovery. Approaches based on CRISPR technology have been broadly adopted and there is growing interest in the generation of massively parallel edited cell libraries. Comparing the libraries generated by these varying approaches is challenging and researchers lack a common framework for defining and assessing the characteristics of these libraries. Here we describe a framework for evaluating massively parallel libraries of edited genomes based on established methods for sampling complex populations. We define specific attributes and metrics that are informative for describing a complex cell library and provide examples for estimating these values. We also connect this analysis to generic phenotyping approaches, using either pooled (typically via a selection assay) or isolate (often referred to as screening) phenotyping approaches. We approach this from the context of creating massively parallel, precisely edited libraries with one edit per cell, though the approach holds for other types of modifications, including libraries containing multiple edits per cell (combinatorial editing). This framework is a critical component for evaluating and comparing new technologies as well as understanding how a massively parallel edited cell library will perform in a given phenotyping approach.

Competing Interest Statement

All authors are or were at one time employees and shareholders of Inscripta, Inc.

Footnotes

* Updated corresponding author.

* https://github.com/InscriptaLabs/cell_lib_eval_paper