What are the limits of spike detectability in GT conditions where we know the waveform(s) we're looking for.
- For a single channel, compute dot product between the template (average spike waveform) of that channel and each of the individual extracellular GT spikes.
- For same channel, compute dot product between the template and segments of the recording.
- Plot the frequency distribution of dot product results in two conditions. Can the distributions be separated?
- Systematically vary threshold used in each case and plot ROC curve. Outlook: Potential for using generative models + template matching for spike detection.
Spike propagation trajectory across multiple channels
Given a probe insertion close to parallel to the somato-dendritic axis of cortex, is it possible to track spike propagation of a unit across channels?
- Work out AP spatial spread along D-V axis
- Map propagation trajectory for different units
- Analyse features of propagation in space
- Tie into sub-cellular AP signatures Outlook: Viability of investigating BAPs in vivo; use of multi-channel EAP propagation stereotipy as an added dimension for sorting.
What else can we tell about them from this dataset?
- Analyse EAP and patch spike waveforms and compare to good units at similar distance range Outlook: Further characterisation of dark neurons and their identity.
How can we take advantage of a paired-recording dataset to explore analyses for connectivity?
- Spike sort recordings where whole-cell patch-clamp was achieved and there was a good EAP signature.
- For every unit in the recording, run STA on the whole-cell unit's membrane potential.
- For every putative pre-synaptic unit, run STA and obtain extracellular footprint; investigate for AxTPs Outlook: Finding additional signatures for unit-unit connectivity.