Explore and Analyze your Recordings Faster!
What kind of database?
2 Solutions are avaible :
A local SQLite database. It's just a file on your hard drive, with all RAW data inside. It's fast and easy to use
A remote MySQL database. It's a little bit slower (because all data have to transit through the local network), but data and analyzes can be shared between users. And backup is easy to perform.
Why to use a database?
3 Main advantages:
First, You can easily tag your traces. It's usefull to excluses noisy traces, artefacts etc... But tags can also contains information about stimulation protocol, used pharmacology etc... Thus, you can quickly find your data, and sort them using filters.
Second, you can do transversal analyses, by concatenating in all your experiments the sweeps that match your wanted parameters.
For example, you can in a few clicks sort all the sweeps that have a leak < to 200 pA, or a spontaneous spiking frequency between 42 and 59Hz, or mean amplitude > 75pA...
Third, You can re-analyse everything in just one click. Because you always start from the RAW data, you can do a whole script that will get your data, analyze the signal and sort the final figure in an EPS file... And this makes data exploration much easier and powerfull. You don't have to deal with all the intermediate data columns in IgorPro, Excel or Sigmaplot, storing.
Let's say you are studying short term plasticity at a given synapse. You have to compare 5 stimulation frequencies, and the best would be to have 20 sweeps for each frequency.
Case 1 (left) ; The easiest way is to do the protocol #1 20 times, then the protocol #2 20 times etc...
Everything is great if you recording is perfect (no changes in R serie, in stimulated cell excitability, no induced long term plasticity...), because you can compare protocol #1 average with all the other averages. But usually, no recording is perfect, and maybe you don't know what will be the effect of 20 successive repetition of protocol #1! You may see some differences but they could be due to other factors.
Case 2 (right): To minimize this risk, the best protocol would be to do protocol #1, then #2, #3, #4, #5, and then back to #1 etc... So if there is a trend, let's say due to an increase in Rserie, all the recordings from the 5 stimulation protocols will be affected equally. Moreover, if you cell dies at recording 80, in case 1 you would only have 20 repetitions of protocols #1, #2, #3 and #4. In case 2, you would have all the 5 protocols, but with a smaller average of only 16 sweeps. Which is enough for a short term plasticity average.
Case 2 is way much better, but will be very hard to analyze because all protocols are mixed. This is where a powerfull Tag system is usefull.
The advantage of multiplexed protocols
