Fig 4

Generalized linear model (GLM) of V1 and PPC responses. A generalized linear model was fit to training data to predict the calcium responses of individual V1 and PPC neurons using lagged predictors for stimulus, engagement and motor action (licking). a GLM model performance for an example V1 neuron. GLM fits on Target (red) and Non-target (blue) trials were compared with trial-averaged test data (black) across Passive (PassNT, PassT), Engaged No-Lick (Miss, FA), and Engaged Lick (FA, Hit) conditions. GLM components related to Stimulus (purple), Engagement (gray), and Motor (green) were summed to generate overall GLM fit. Shaded region represents time of stimulus presentation. b Same as a, but for a different, nontarget-preferring V1 neuron. c, d GLM model performance for two example PPC neurons. One neuron has a strong Engagement component (c), while the other has a stronger Motor component (d). e Cumulative histogram of model prediction performance on test trials, quantified for each cell as fraction of variance explained (R²), for both V1 (blue, n = 1915 cells) and PPC (red, n = 3524). f Population-averaged GLM model components related to Stimulus (left), Engagement (middle), and Motor (right), for both V1 (blue) and PPC (red). For Stimulus and Engagement components, the preferred stimulus component for each cell was used for averaging. Calcium responses were z-scored before model fitting. Light shaded gray region demarcates duration of stimulus