This article presents a scholarly dispute about cell growth patterns in Schizosaccharomyces pombe (fission yeast). Cooper argues that published cell length data, originally interpreted as showing two linear growth segments separated by a rate-change point (RCP), are equally well-explained by exponential growth. Using semi-logarithmic plots and statistical regression analysis (r² values), Cooper demonstrates that a single exponential model with two parameters fits the data nearly as well as a three-parameter bilinear model. Applying Occam's Razor, he advocates for the simpler exponential explanation. Mitchison and colleagues counter that Cooper's analysis ignores marked, non-exponential features of yeast growth: length extension ceases in the final ~20% of the cell cycle, and rate analysis using more sensitive difference-curve methods reveals clear bilinear patterns with ~30% rate changes during interphase and substantial rate changes at cell division. They argue that different growth parameters (volume, mass, protein) show different patterns, none simply exponential, and that wee1 mutants show even more conspicuous rate changes incompatible with exponential growth.

Key findings

• Statistical r² values for exponential versus bilinear models are negligible, making model choice depend on parsimony rather than fit quality
• Semi-logarithmic plotting versus rate difference analysis yield different interpretations of the same fission yeast length data
• Cell growth in S. pombe involves multiple rate-change points including cessation of length extension and changes at division that cannot be explained by simple exponential kinetics
• Sensitivity of analytical methods matters critically: smoothed rate-difference analysis reveals bilinear growth patterns invisible to simple regression approaches
• Different growth parameters (length, volume, mass, protein) exhibit different kinetics, none fully exponential