Abstract
Neurophysiological studies depend on a reliable quantification of whether and when a neuron responds to stimulation. Simple methods to determine responsiveness require arbitrary parameter choices, such as binning size, while more advanced model-based methods require fitting and hyperparameter tuning. These parameter choices can change the results, which invites bad statistical practice and reduces the replicability. New recording techniques that yield increasingly large numbers of cells would benefit from a test for cell-inclusion that requires no manual curation. Here, we present the parameter-free ZETA-test, which outperforms t-tests, ANOVAs, and renewal-process-based methods by including more cells at a similar false-positive rate. We show that our procedure works across brain regions and recording techniques, including calcium imaging and Neuropixels data. Furthermore, in illustration of the method, we show in mouse visual cortex that 1) visuomotor-mismatch and spatial location are encoded by different neuronal subpopulations; and 2) optogenetic stimulation of VIP cells leads to early inhibition and subsequent disinhibition.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
In addition to many small changes, we made the following four major additions: 1) We added a new analysis showing that the ZETA-test displays a sensitivity close to that of a t-test to cells with purely exponentially-distributed inter-spike intervals (ISI), as well as a superior sensitivity to cells with no peaks in activity, but with highly-peaked ISI distributions, such as bursting cells (Figure 4; the old figure 4 is now Figure 3 - Supplement 1). 2)We added a comparison with other sophisticated, state-of-the-art methods, such as the multiplicative inhomogeneous Markov interval (MIMI) model (Figure 3 - Supplement 2, Figure 6). These comparisons show that the ZETA-test and ZETA-IFR outperform MIMI-based approaches respectively for responsiveness detection and latency determination. 3) We added a much more detailed mathematical description of the ZETA-test's components, and why they are important, to the method section. This includes a comparison of the ZETA-test with alternative formulations derived from the starting point of renewal-process models (figure 3 - supplement 3). Moreover, the supplementary methods include the derivation of closed-form solutions of some of ZETA's important intermediate variables which highlight the difference between the Kolmogorov-Smirnov test and the ZETA-test. 4) We have expanded the discussion to include a section on the scope and applicability of the ZETA-test, highlighting its use as an unsupervised generalist statistical test rather than as an alternative to model-based approaches applied to single (classes of) cells.