Abstract
Microbial fitness screens are a key technique in functional genomics. We present an all-in-one solution, pyphe, for automating and improving data analysis pipelines associated with large-scale fitness screens, including image acquisition and quantification, data normalisation, and statistical analysis. Pyphe is versatile and processes fitness data from colony sizes, viability scores from phloxine B staining or colony growth curves, all obtained with inexpensive transilluminating flatbed scanners. We apply pyphe to show that the fitness information contained in late endpoint measurements of colony sizes is similar to maximum growth slopes from time series. We phenotype gene-deletion strains of fission yeast in 59,350 individual fitness assays in 70 conditions, revealing that colony size and viability provide complementary, independent information. Viability scores obtained from quantifying the redness of phloxine-stained colonies accurately reflect the fraction of live cells within colonies. Pyphe is user-friendly, open-source and fully documented, illustrated by applications to diverse fitness analysis scenarios.
Highlights
Colony fitness screens are a key assay in microbial genetics
High-throughput colony-based screening is a powerful tool for microbiological discovery and functional genomics
Using a set of diverse wild yeast strains, we show that the fitness correction approach implemented in pyphe effectively reduces noise in the data
Summary
Colony fitness screens are a key assay in microbial genetics. The availability of knock-out libraries has revolutionised reverse genetics and enabled the field of functional genomics (Giaever and Nislow, 2014). Highly precise fitness determination has been achieved by high-resolution, transilluminating time course imaging and growth curve analysis (Takeuchi et al, 2014) and combined with a reference grid normalisation (Zackrisson et al, 2016). The parallel use of commercially available scanners, combined with high-density arrays of colonies can enable growth curve-based phenotyping at very large scales, but poses challenges in terms of data storage, processing, equipment and the need for temperature-controlled space. We investigate the relationship between colony sizes and viability scores in a broad panel of S. pombe knock-out strains in over 40 conditions and find that the two approaches provide orthogonal and independent information. We link colony redness scores to the percentage of dead cells in a colony and show that phloxine B staining provides similar results as a different live/dead stain
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
