Quantitative Fitness Analysis shows that NMD proteins and many other protein complexes suppress or enhance distinct telomere cap defects in budding yeast

Stephen G. Addinall*, Eva Holstein*, Conor Lawless*, Min Yu, Kaye Chapman, A. Peter Banks, Hien-Ping Ngo, Laura Maringele, Morgan Taschuk, Alexander Young, Adam Ciesiolka, Allyson Lister, Anil Wipat, Darren J Wilkinson and David Lydall

* These authors contributed equally to this work

Supporting Information Data Files

Manuscript (Open Access)

The full manuscript is freely available for download here.

General Information

4400 yeast array deletions were crossed to a ura3::NATMX (controlQFA, or cQFA, DLY4228), yku70::URA3 (DLY3541) or cdc13-1 strains (DLY5385) as described in supplementary material. Double mutant cultures were incubated at different temperatures and photographs taken every few hours. Strains were tracked using barcoded plates and a Robot Object Database (ROD). Growth was quantified using an image analysis tool Colonyzer (Lawless et al. 2010). The range of observations were summarised by fitting a logistic growth model by least squares optimisation (see Fig. 1 (b)). Fitnesses were calculated and genetic interaction strengths (GIS) were inferred by comparing with a control QFA (see Fig. 2). GIS profiles were compared across different experiments (see Fig. 3)

Hitlists (Supplementary Tables S1-S6 & S9)

These are tables, hyperlinked to gene entries in SGD, and tab-delimited text files containing estimates of genetic interaction strength and significance as calculated by comparing double mutant fitness phenotype against a control QFA fitness phenotype. Only mutants with GIS which pass the FDR-corrected significance test (q-value<0.05) are included.

Strong lists (S1-S6) only include array mutants with GIS < -0.5 and GIS > 0.5 and with q-value < 0.05, and are presented in the supplementary material of the manuscript. A GIS cutoff of 0.5 is arbitrary. Full lists (text files) and hitplots are not included in the supplementary material of the manuscript but include all interactions with q-value < 0.05. Hitplots are visualisations of the genetic interactions quantified in these lists, equivalent to of those presented in Fig. 2. Points in hitplots represent mean fitness (over ~8 biological repeats) in double-mutant arrays and in single-mutant temperature-matched control arrays. Solid grey line represents regression through observations, dashed line is the 1:1 line. See Fig. 2 for further details.

Raw data files

These files contain quantified cell densities output from our ROD database. There are eight tab-delimited text files (one for each biological repeat) containing various measures of culture density, colour and morphology (as estimated by Colonyzer for each mutant in the experimental library.

Each column represents the following:

Inoculation dates for each of the screens. Screen names correspond to those in raw data files above and logistic data files below. Inoculation date-time specified as YYYY-MM-DD_hh-mm-ss.

Logistic data files

These files contain logistic parameter estimates, summarising timeseries observations of culture density from the ROD output above. There are eight tab-delimited text files (one for each biological repeat) containing logistic parameter estimates for various measures of culture density, together with estimates of final culture colour and morphology for each mutant in the experimental library.

Detailed instructions describing how we converted these logistic parameter estimates to fitness estimates are included in the Quantifying Fitness subsection of the Experimental Procedures section of the manuscript.

Each column represents the following:

Temperature matches used to compare between cQFA and double mutant fitness phenotypes. This is a tab-delimited text file specifying query mutation, double mutant treatment, and neutral query mutant (cQFA) treatment for each genetic interaction scoring comparison.