You can print this document to have a physical copy to hand while using iRVis.
By comparing the fitnesses of independent microbial strains grown under two different environmental or genetic conditions (e.g. query and control conditions) we can rank and compare the strength of gene-gene or gene-environment interactions. Several technologies for measuring microbial fitness exist, including Quantitative Fitness Analysis (QFA), Synthetic Genetic Array (SGA) and liquid culture growth assays. Genome-wide fitness comparisons can become difficult to visualise by static, 2D scatterplot due to the sheer number of strains examined because simultaneous labelling of ~4,000 genes on a single plot is not practical. For example, in Figure 1, the labels for many genes are obscured.
Figure 1: A static fitness plot Fitness plot showing evidence for gene-gene interactions between a query mutation and each of the deletions in the yeast knock-out collection. Mean fitnesses were calculated for each deletion combined with a temperature sensitive query mutation (yku70Δ) and plotted on the y-axis. Mean fitnesses were calculated for each deletion combined with a neutral control mutation (ura3Δ) and plotted on the x-axis. The screens were approximately temperature matched (~37°C). Red and green points indicate genotypes which significantly suppress and enhance the fitness defect of the query mutation, respectively. Blue horizontal and vertical lines intersect at the point corresponding to his3Δ (a wild-type surrogate). Solid grey line is predicted double-mutant fitness, given single deletion fitness and assuming a multiplicative model of genetic independence. Dashed grey line is the line of equal fitness.. This static fitness plot was generated using an early version of the QFA R package and is a reproduction of Figure 2 from Addinall et al., 2011
iRVis is a tool for generating interactive versions of fitness plots (as shown in Figure 1). iRVis allows us to make rapid, visual comparisons between different pairs of QFA experiments and to query plots in real time, aiding analysis and interpretation of the underlying data. iRVis can be applied to any paired sets of control and query fitnesses, including SGA data, or fitnesses derived from liquid growth curves.
We are also currently developing DIXY, a web-based alternative with broadly similar functionality. DIXY has the advantage that it requires no installation, it simply runs in your browser. However, transitions between plots currently run much slower using the web-based tool and so, for the moment, we recommend the R-based version for browsing multiple datasets.