Feel like these kind of vigorous benchmarks are not getting enough attention among all the AIVC talk. "The central contribution of our note is to recognize that perturbation prediction metrics are implicitly point estimates of population parameters."
Rafael Irizarry@rafalab.bsky.social · 3mo agoA flaw in Person-Δ may be overstating progress in single-cell perturbation prediction models. Pearson warned about this in the 19th century: reusing the same controls induces spurious correlation. Split the controls, and much of the claimed prediction power fades: www.biorxiv.org/content/10.6...
