Good Day.
I am doing patch clamp recordings on neurons and HEK cells. I would like to know what are the recommended sampling rates for neurons and HEK cells. I am concerned that my high sampling rate for HEK cells (~10 KHz), will occupy my disk space. If I reduce my sampling rate, I am concerned with possible changes in my recording trace.
In your case, what is your suggested sampling rates of each cell? Is there a reference or basis for choosing these suggesting sampling rates?
Hoping for your insightful answers.
Thanks
I think 10 kHz is a good sampling rate for neuronal recordings. Action potentials are quite fast -- and some of the underlying currents are even faster. I don't think you'd ever want to filter any lower than 3 kHz, which means by the Nyquist theorem that 6 kHz is the lowest sampling rate permissible. I usually use 20 kHz in current clamp and, when trying to isolate especially fast currents in voltage clamp, 50 kHz or more.
You probably would be able to go slower with HEK cells, depending on what the cells were expressing.
But I'm not sure why you're worried about disk space. At double precision, a single sample point uses 8 bytes. Say that you patch onto a cell and administer a 1-sec-long step 100 times. So, 100 seconds of recording per cell. That requires 8 bytes/sample * 10000 samples/sec * 100 sec = 8 MB*. If this were 1988, your floppy disk would be overwhelmed, but in 2023 this is nothing. You can buy a 32 GB USB flash drive on Amazon for $6 -- enough for 4000 recordings!
Some things (imaging, Neuropixels) require a lot of space, but patch clamp does not. I don't think you need to worry about this.
*In this calculation, I assumed 1024 is basically the same as 1000.
Dear Simon Miguel Lopez ,
Niraj S. Desai explained beautifully the idea of sampling rate vs disk space. I'll only add my own experience here. I had a Mac for 5 years which I used for recordings. I sampled everything at 100 kHz - I'm always concerned with currents and AP kinetics - and I never reached 1 TB. I dare to say I never filled up 500 GB. If you're doing 10 or 20 kHz for room temp currents, you should be fine. But keep in mind that if you go physiological temperature or want to dig into AP waveform, crank it up to 50 kHz at least.
Good luck,
Andre
I just like to add that the cell type is not what matters when chhosing a sampling rate but rather the type of ion channel (fast kinetics of activation vs. slow) and the patch clamp configuration mode (patch vs. whole cell).
- Badges
- Science method