This README file describes the `v2` exported properties for the MICrONS digital twin. # Contents Properties are located in the following subdirectories: - readout - performance - ori_dir_tuning - responses - anatomy # Subdirectory contents ## readout This subdirectory contains readout weights and readout locations for the digital twin model units. ``` NAME SHAPE ROWS PROPERTY readout_weights.npy 104171 x 512 units readout_weight_vector readout_locations.npy 104171 x 2 units readout_location_x, readout_location_y units.csv 104171 x 3 units session, scan_idx, unit_id metadata.csv 13 x 6 scans animal_id, session, scan_idx, ... ``` ## performance This subdirectory contains performance metrics for the digital twin model units. ``` NAME SHAPE ROWS PROPERTY units.csv 104171 x 6 units session, scan_idx, unit_id, cc_abs, cc_max, cc_norm metadata.csv 13 x 11 scans animal_id, session, scan_idx, ... ``` ## ori_dir_tuning This subdirectory contains orientation/ direction tuning metrics for the digital twin model units. Note: These properties are from a slightly different version of the digital twin that had additional training on oriented stimuli. See methods in [2]. ``` NAME SHAPE ROWS PROPERTY units.csv 104171 x 9 units session, scan_idx, unit_id, OSI, DSI, gOSI, gDSI, pref_ori, pref_dir metadata.csv 13 x 10 scans animal_id, session, scan_idx, ... ``` ## responses This subdirectory contains computed responses for the digital twin model units and the stimulus used to generate them. ``` NAME SHAPE ROWS PROPERTY responses.npy 104171 x 75000 units response stimulus.npy 75000 x 144 x 256 frames image (height x width) units.csv 104171 x 3 units session, scan_idx, unit_id metadata.csv 13 x 11 scans animal_id, session, scan_idx, sampling_frequency, ... ``` ## anatomy This subdirectory contains the anatomical information for the digital twin model units. ``` NAME SHAPE ROWS PROPERTY units.csv 104171 x 8 units session, scan_idx, unit_id, unit_x, unit_y, unit_z, brain_area, field metadata.csv 13 x 7 scans animal_id, session, scan_idx, ... ``` # Property descriptions For more details on each property see methods section of corresponding citation. Citations are listed in the section below. ``` NAME DESCRIPTION CITATION animal_id ID of the mouse [1 - 3] session ID of the imaging session [1 - 3] scan_idx ID of the scan within the imaging session [1 - 3] unit_id ID of the unit within the scan [1 - 3] readout_weight_vector Readout weights for the unit learned by digital twin model [1, 2] readout_location_x X coordinate of the readout location, an approximation of receptive field center in stimulus space; (-1, -1) top-left, (1, 1) bottom-right [1, 2] readout_location_y Y coordinate of the readout location, an approximation of receptive field center in stimulus space; (-1, -1) top-left, (1, 1) bottom-right [1, 2] cc_abs Test set performance of the digital twin model unit, higher is better [1, 2] cc_max Neuron variability score used to normalize digital twin model unit performance [1, 2] cc_norm Normalized model unit performance, higher is better [1, 2] OSI Orientation selectivity index [2] DSI Direction selectivity index [2] gOSI Global orientation selectivity index [1, 2] gDSI Global direction selectivity index [1, 2] pref_ori Preferred orientation in degrees (0 - 180), vertical bar moving right is 0 and orientation increases counter-clockwise [1, 2] pref_dir Preferred direction in degrees (0 - 360), vertical bar moving right is 0 and orientation increases counter-clockwise [1, 2] response Response for each unit to a stimulus frame [1, 2] image Stimulus image (height x width) [1, 2] sampling_frequency Frequency used to resample stimulus video shown to model [1] unit_x X-coordinate of unit in cortical space (posterior -> anterior) [3] unit_y Y-coordinate of unit in cortical space (lateral -> medial) [3] unit_z Z-coordinate of unit in cortical space (superficial -> deep) [3] brain_area Visual cortical area label [3] field ID of the two-photon imaging plane, given animal_id, session, scan_idx [3] ``` # Citations ``` ID NAME DOI 1 Wang et al., 2025 https://doi.org/10.1038/s41586-025-08829-y 2 Ding, Fahey, and Papadopoulos et al., 2025 https://doi.org/10.1038/s41586-025-08840-3 3 MICrONS Consortium et al., 2025 https://doi.org/10.1038/s41586-025-08790-w For use of the digital twin and any properties please cite [1, 2] For use of the anatomical properties please cite [3] ```