A factorial design was used for estimating variability due to Chip, Species and Kit, consisting initially of 8 sequencing runs (Table S2). A second experiment using a randomised Plackatt-Burman design consisting of 4 runs was conducted to help estimate inter-machine variability (Table S3). However, the Ion Xpress Template 200 kit was phased out during the experiment, preventing the completion of the final experiment. The release of a new kit part-way through the data-generation necessitated the addition of four new datasets, also conducted using a Plackatt-Burman design (Table S4). In total, 15 datasets were generated.
