Fitness impact of N- versus C-terminal tagging. (a) Evaluation of the SWAT tagging efficiency during construction of the Halo-ORF and ORF-Halo libraries using donors with mNG as the tag. Tagging was performed with 13 N-SWAT (left) or C-SWAT (right) strains for the indicated ORFs. Distributions of single-cell mNG fluorescence intensities measured with flow cytometry (10,000 cells per strain). The percentage of cells with fluorescence above the background (fluorescence of a wild type [wt] strain, dashed line) is indicated. (b) Reproducibility of relative fitness estimation with the competition assay. Relative fitness of two replicates for 92 ORFs, determined as in Fig. 1 a. r, Pearson correlation coefficient. (c) Evaluation of technical noise in the competition assay. Distributions of relative fitness determined as in Fig. 1 a for the indicated ORFs and the pair of SWAT donor strains (no tagged ORF), 92 technical replicates per strain pair. Mean ± SD relative fitness is indicated. (d and e) Triple competition of 45 randomly selected pairs of Halo-ORF and ORF-Halo strains against a wild type strain lacking fluorescent markers. Relative fitness of Halo-tagged strains determined according to Fig. 1 a and in the triple competition assay (d, Materials and methods). Pairwise relative fitness of the three strains for each ORF in the triple competition assay (e, Materials and methods). Median relative fitness (0.13) of N- and C-terminal variants against wild type (dashed line, e). (f) Proteins with localization signals at the N-terminus, C-terminus, or both protein termini, stratified by gene essentiality and differential fitness according to Fig. 1 b. Analysis based on 997 validated or predicted localization signals: 611, 188, and 99 proteins with signals at the N-terminus, the C-terminus, or at both termini, respectively (Materials and methods, Table S2). N-loser, neutral, and C-loser ORFs are defined at two relative fitness thresholds.