%0 Journal Article %A Ludovica Bachschmid-Romano %A Nicholas G. Hatsopoulos %A Nicolas Brunel %T Interplay between external inputs and recurrent dynamics during movement preparation and execution in a network model of motor cortex %D 2022 %R 10.1101/2022.02.19.481140 %J bioRxiv %P 2022.02.19.481140 %X The primary motor cortex has been shown to coordinate movement preparation and execution through computations in approximately orthogonal subspaces. The underlying network mechanisms, and in particular the roles played by external and recurrent connectivity, are central open questions that need to be answered to understand the neural substrates of motor control. We develop a recurrent neural network model that recapitulates the temporal evolution of single-unit activity recorded from M1 of a macaque monkey during an instructed delayed-reach task. We explore the hypothesis that the observed dynamics of neural covariation with the direction of motion emerges from a synaptic connectivity structure that depends on the preferred directions of neurons in both preparatory and movement-related epochs. We constrain the strength both of synaptic connectivity and of external input parameters by using the data as well as an external input minimization cost. Our analysis suggests that the observed patterns of covariance are shaped by external inputs that are tuned to neurons’ preferred directions during movement preparation, and they are dominated by strong direction-specific recurrent connectivity during movement execution, in agreement with recent experimental findings on the relationship between motor–cortical and motor–thalamic activity, both before and during movement execution. We also demonstrate that the manner in which single-neuron tuning properties rearrange over time can explain the level of orthogonality of preparatory and movement-related subspaces. We predict that the level of orthogonality is small enough to prevent premature movement initiation during movement preparation; however, it is not zero, which allows the network to encode a stable direction of motion at the population level without direction-specific external inputs during movement execution.Competing Interest StatementThe authors have declared no competing interest. %U https://www.biorxiv.org/content/biorxiv/early/2022/02/19/2022.02.19.481140.full.pdf