Analysis code and spike count data for our paper State-dependent geometry of population activity in rat auditory cortex (eLife, 2019).
- Paper: https://elifesciences.org/articles/44526
- Preprint: https://www.biorxiv.org/content/10.1101/501338v2
The spike count data are located in the spicounts folder. Each file corresponds to one experimental session. Columns correspond to neurons and the first row encodes the shank from which this neuron was recorded. The first six columns of the remaining rows are: (1) session id after splitting some of the sessions (some experimental sessions were split in two based on the activation level, see paper); (2) CV of the [splitted] session; (3) trial number; (4) ABL; (5) ILD; (6) center of the 50 ms spike count window relative to the sound onset.
Run preprocessSpikecounts.m to produce evokedResponses.mat that is used by all other scripts. The preprocess.m script loads the original spiking data (not shared here) and also produces evokedResponses.mat.
The decoding analysis is done in the runDecoding.m and takes some hours. The resulting performance metrics are saved into decoding_results.mat directly available in this repository.
Each figure is produced by the corresponding figure[...].m script.