Downloading/Accessing Electrophysiology Data for Lee et al., 2023

There is a new paper published this year, and I have found the SWC files for morphologies on GitHub. But I cannot find a way to access the electrophysiology sweeps (I used to access the Cell Type Database with AllenSDK, but I assume I can figure something out if someone can point me to the NWB files for the new study as well). I am interested in finding out the stimulus duration and number of spikes for the sweep right above rheobase for each neuron (the “representative sweep”).

Hi, thank you for your interest in the paper. The nwb files can be found on DANDI (DANDI Archive). We also have a github location for our data (GitHub - AllenInstitute/human_patchseq_gaba). Please let me know if you have any additional questions.

1 Like

Thanks Brian!

I checked the DANDI page, where nwb files are group in folders with names like “sub-1000181910”. I assumed that the numeric string after the “sub-” is the specimen ID? However, I cannot find matching IDs with the swc files nor the values in the specimen_id column of the LeeDalley_ephys_fx.csv file on github.

How can I get the corresponding nwb files for a given specimen id?

Additionally, for each nwb file, when I open them in HDFview, I see the “subject_id” under “subject”, but no cell IDs/specimen IDs.

@brian_lee Sorry, replied to myself instead of replying to you.

With the supplied asset file on DANDI, it also does not list the specimen IDs. In the original Allen Cell Type Database, there is a json file for all cells that looks like this that does include a “specimen__id” for each cell (I am including only one cell here): [ { "cell_reporter_status": null, "csl__normalized_depth": null, "csl__x": 273.0, "csl__y": 354.0, "csl__z": 216.0, "donor__age": "25 yrs", "donor__disease_state": "epilepsy", "donor__id": 524848408, "donor__name": "H16.03.003", "donor__race": "White or Caucasian", "donor__sex": "Male", "donor__species": "Homo Sapiens", "donor__years_of_seizure_history": "9", "ef__adaptation": 0.0278459596639436, "ef__avg_firing_rate": 13.5725111407696, "ef__avg_isi": 73.6783333333333, "ef__f_i_curve_slope": 0.1671875, "ef__fast_trough_v_long_square": -53.8750038146973, "ef__peak_t_ramp": 4.10410666666667, "ef__ri": 159.531131386757, "ef__tau": 21.1810256736186, "ef__threshold_i_long_square": 90.0, "ef__upstroke_downstroke_ratio_long_square": 2.89546090494073, "ef__vrest": -70.56103515625, "ephys_inst_thresh_thumb_path": "/api/v2/well_known_file_download/529903142", "ephys_thumb_path": "/api/v2/well_known_file_download/529903140", "erwkf__id": 618211597, "line_name": "", "m__biophys": 0, "m__biophys_all_active": 0, "m__biophys_perisomatic": 0, "m__glif": 5, "morph_thumb_path": null, "nr__average_contraction": null, "nr__average_parent_daughter_ratio": null, "nr__max_euclidean_distance": null, "nr__number_bifurcations": null, "nr__number_stems": null, "nr__reconstruction_type": null, "nrwkf__id": null, "si__height": null, "si__path": null, "si__width": null, "specimen__hemisphere": "right", "specimen__id": 525011903, "specimen__name": "H16.", "structure__acronym": "FroL", "structure__id": 12113, "structure__layer": "3", "structure__name": "\"frontal lobe\"", "structure_parent__acronym": "FroL", "structure_parent__id": 12113, "tag__apical": "intact", "tag__dendrite_type": "spiny" },...]

hi @Veritatem_Amo, I have added a new metadata file to our github page (LeeDalley_manuscript_metadata_v2.csv). This sheet has the NWB file name which you will be able to associate the files from the DANDI link.

1 Like

Thank you very much! This is exactly what I need! @brian_lee

Just to make sure before I start my downstream analysis, there are about 16 NWB files from DANDI that do not have a specimen ID based on the metadata file. Also, there are 44 NWB IDs from the metadata file that does not have a NWB file from DANDI.

Does this sound about right?

@brian_lee One additional question, in the metadata for electrophysiology, is “v_baseline” the resting membrane potential? To my understanding, “rheo” in the column names is short for rheobase; what is “hero” in the column names?

Yes v_baseline is resting membrane potential and rheo is short for rheobase. Hero refers to a suprathreshold voltage trace that contains spiking.

Hi @brian_lee,

For the morphology of a lot of the neurons from Brain Image Library, there are two SWC files, one ends with “_transformed.swc”, the other ends with “_raw.swc”. In some cases, instead of a “_transformed.swc” file, there is a “_upright.swc” file. Sometimes there is a single file ending with "_m.swc. What are the distinctions in between?

If I want to model surface area in μm², for instance, which file should I use (right now _transformed.swc files and _raw.swc files of the same cell give different numbers)?

Thanks and happy new year!