tom_olfactometer_configs: A repository from tom-f-oconnell

Current protocol

Initial setup

IMPORTANT: if another olfactometer is connected and you need to take it off, power off the narrow left power strip (which currently has the power supply for Dhruv's olfactometer connected) before disconnecting the valves.
Connect the outputs of the two 200 mL/min flow meters to the two 12-valve manifolds on my olfactometer (follow color codes). Connect the 1600 mL/min flow meter to the PTFE tube that will accept the outputs of the odor vials.
Connect the two D-sub connectors that deliver power to the valves. Each manifold gets exactly one connector. There is yellow tape in three places: on the end of one of the D-sub cables, on the only connector that doesn't have red tape over it for the back manifold, and on one of the two D-Sub connectors mounted to the front/left manifold. Connect the cable with yellow tape to the connector from the back manifold that also has yellow tape. Connect the cable without yellow tape to the D-sub connector on the left manifold that also does not have yellow tape on it.
Connect the two BNC cables connected to my Arduino (each should currently be labelled with yellow or green tape and have some kind of BNC adapter on the end) to the corresponding cables going to the ThorLabs data acquisition hardware (which should be color coded in the same manner with yellow and green tape).
Disconnect the USB cable for the Arduino in Dhruv's olfactometer and connect the USB cable for the Arduino in mine (should have labelled yellow tape in it).
Check that the valves between the wall vacuum valve and the vacuum funnel inside the dark box are allowing flow through the funnel, though when you set up a fly later check that there is still sufficient suction to evacuate the excess saline. I stick my finger in the funnel to form a seal and see that it has noticeable suction within a second or so.
Remove the PID head from the dark box if it's in there.
Open ThorSync / ThorImage version 3.0 (the ones pinned to the taskbar) (and all other necessary software).
Check that flow meters are set to 1600/200/200 mL/min. People who do experiments only using one manifold will typically have the higher flow rate set to 1800 mL, and typically this is all that needs to be changed.

Test the valves

Open a Windows command prompt (the black icon on the task bar). You can use this same command prompt for all of the subsequent steps involving valves control / odor delivery.
```
cd src
cd tom_olfactometer_configs
```
Run the valve test program and listen for clicks for each valve.
```
olf-test-valves -u
```
The valve numbers get printed to the terminal as the program executes, and each valve should be switched three consecutive times by default.

The -u just re-uploads the firmware my code expects to the Arduino, so that if someone else was using their own Arduino scripts, it will still work. No subsequent commands will require this once you do it once. If you forget, the command will fail with some error.

For each fly

If you ever want to stop stimulus delivery: go to the windows command prompt running the stimulus delivery, press Ctrl-C, and then run olf-upload. If you don't run olf-upload, the Arduino will keep running the rest of the stimulus delivery program, even though nothing would be printed to the terminal.

Create a directory D:\Tom\<YYYY-MM-DD>\<number of the current fly, starting at 1>
- Make sure ThorSync is saving to this numbered directory. Start the ThorSync directory name at SyncData001, and you should be able to let it autoincrement from there within each fly. For each new numbered fly directory, start the ThorSync output directory name at SyncData001 again.
Glomeruli diagnostic panel
- In the Windows command prompt (that is still in the src\tom_olfactometer_configs directory):
```
olf glomeruli_diagnostics.yaml
```
  It will print which odors you should connect to which valves, and will wait for you to press Enter before it starts the experiment. For each manifold, you will be exclusively using the valves on the on the top row (this will mean the front manifold will have the higher valve numbers and the back manifold will have the lower valve numbers).
  
  Note that there should be one solvent vial connected to each of the valves with a white tube downstream of it (there is one such valve per manifold). There should always be solvent vials connected to these two valves, though the program does not tell you to connect solvent here, as there should always be solvent connected downstream of these valves.
  
  First load the odor vials in to one of the 6-vial holders, then press each of the holders in one of the pairs of brackets on either side of the PTFE tube. Try to order the vials in the holders / angle them such that no tubing gets kinked or strained too much.
  
  Wait to press Enter until after you have the fly on and are done with the ThorImage setup described below.
  
  Check that all three of the mass flow controllers are achieving their set points.
- ThorImage setup:
  - 192 x 192 pixels (drag the slider to get the values not available in the drop down)
  - ~5.9 - 6.3ish zoom. Pick a value when stepping with the piezo (as described below) such that there is at least a roughly 5% margin on all sides of the antennal lobe in case there is some slight drift.
  - Initial plane should be just low enough that it's still roughly circular and it's not just a couple glomeruli in plane. Set piezo step size to 12um in the live mode and step through from 0 to 48 um to make sure that the antennal lobe stays in the field of view in each plane.
  - ~2.5 power, though I was using ~3-4 before when it seems the scope power might have been slightly lower. Shouldn't matter too much.
  - Either antennal lobe should be fine (fly should be mounted on scope so it is facing you). Just pick the better looking one. If one has more inhomogenous baseline fluorescence (some bright glomeruli), and all other things look equally good, go for the side without this kind of elevated baseline.
  - (now switch to the capture plane) Streaming -> Stimulus
  - Raw data (not TIFF) output
  - Enable fast Z
  - Fast Z parameters: start=0, stop=48, step=12 (all same as when testing in live)
  - 1 flyback frame
  - Make sure the data is being saved to the numbered directory created before
  - Name this experiment glomeruli_diagnostics
  - When you ran olf glomeruli_diagostics.yaml above, it should have automatically copied the name of a .yaml file it generated to your clipboard. Click into the ThorImage experiment notes and paste the name of this generated file. If for some reason it's no longer in your clipboard, the terminal running the stimulus delivery will have the name of this file among the output it has printed so far.
    
    Also in the note field, include:
```
odors: <month>/<day> (when the odors were prepared. labelled on vials.)
surgeon: <your-name>
```
  - Make sure ThorSync is connected.
- Make sure the curtain is closed as much as it can be. There should be a fair bit of overlap between the two sections of the curtain.
- Turn the sensitivity on the manual manipulator all the way down and disable as many of the axes as you can.
- Start ThorImage acquisition (and make sure ThorSync is also recording). Now switch back to the terminal and press Enter to start stimulus delivery.
Kiwi approximation components and mix ramp
```
olf kiwi.yaml
```
Set up ThorImage as before. If necessary, make slight adjustments to make the initial plane look as it did in the previous experiment. Step through planes again on the piezo to check that the antennal lobe is still within the field of view.

Start ThorImage, check ThorSync is started, and then press Enter in the terminal to start the experiment.
Ethyl acetate and ethyl butyrate ramps
```
olf kiwi_ea_eb_only.yaml
```
Otherwise follow instructions as in step above.
"Anatomical" Z-stack to assist in glomeruli identifaction / registration

The ThorImage setup is similar here, but with a few important differences:
- Turn on the red channel, by turning the gain of the second PMT to a similar value to the gain of the green channel PMT (~10, or at least no more than 15).
  
  Also click the box (labelled "B", I think?) under the green box (labelled "A") on the right.
- Set the resolution to 384x384 (384 = 192*2, to make for easy downsampling)
- Set the intial plane higher, such that it just barely gets the top of the antennal lobe, rather than trying to capture much signal in the initial plane itself.
- Capture settings:
  - Change mode from Streaming to Z/T-series mode
  - In Z-stack settings: step=0.5um, stop=96um
  - Enable the setting to allow streaming some number of planes at each Z depth, and configure it to stream 5 frames at each depth. This is to try to be able to filter out some motion.
  - Uncheck the time series streaming box beneath the Z-stack section.
  - Disconnect ThorSync (or just delete the data later)
  - Name the experiment anat

Installation

Install my olfactometer repo , following instructions in linked README.
Clone this repo.

git clone https://github.com/tom-f-oconnell/tom_olfactometer_configs

Set the environment variable OLFACTOMETER_HARDWARE_DIR to <where-you-cloned-this-repo>/hardware. On Linux, you can add a line like:
```
# (modify path to point to where you cloned this repo)
export OLFACTOMETER_HARDWARE_DIR="$HOME/src/tom_olfactometer_configs/hardware"
```
...to the end of your ~/.bashrc.

For how to do this on Windows, see instructions in the link to my olfactometer repo above.
Using the same method as in step 3 above, set OLFACTOMETER_DEFAULT_HARDWARE to the prefix of the file under the hardware directory that you want to use.

For example, to use ./hardware/single_manifold2.yaml as the default hardware definition, set this variable to single_manifold2.

See link to olfactometer repo above for commands to run the olfactometer or test the valves.