Protocols — the p0–p3 series
The course experiments are numbered p0 → p3. Each one is a self-contained Arena Studio protocol; most come in a short and a full version.
Each protocol asks a different behavioral question. P0 is mainly for learning the system and checking optogenetic intensity. P1 asks how visual motion and looming stimuli drive behavior. P2 asks how flies control or choose visual objects, especially during optogenetic activation. P3 is reserved for a bonus or conditioning experiment.
Workflow: run the short version first to check that the fly, tracking, visual display, and optogenetic timing make sense. If it looks good, run the matching full version. If short and full were run on the same fly, the analysis can pool them.
P0 is the intro/calibration protocol and is worth running once per line when possible. P1 and P2 are the core student protocols. P3 is the Heisenberg flight-simulator tribute: five named short/full protocol pairs use phase-paired visual stimuli and fixed −1.8 closed-loop coupling.
Design principle: internal comparisons
These protocols are built around comparisons that can be made within the same fly and the same run. Genetic controls matter, but stimulus controls matter too: we want to know whether a response depends on a specific feature of the stimulus, not just on “something changed on the display.”
Examples:
- P0 compares the same visual trials across sham, low-light, and high-light optogenetic blocks.
- P1 pairs opposite grating directions, matched temporal frequencies, loom speeds, loom positions, and looming control stimuli.
- P2 balances object-choice stimuli across left/right positions and compares no-opto versus opto phases within the same behavioral structure.
The analysis will keep these matched comparisons visible instead of treating each trial type as an isolated condition.
| # | Name | What it probes | Closed-loop / FicTrac? | Short | Full |
|---|---|---|---|---|---|
| p0 | Optogenetic intensity | Which LED level drives the fly | No (open-loop) | about 2.6 min | about 10.3 min |
| p1 | Visual motion | Optomotor turning + looming response | No (open-loop) | about 2.7 min | about 7.9 min |
| p2 | Object responses | Bar fixation + A/B choice | Yes (needs FicTrac) | about 4.3-4.5 min | about 9.8-10.2 min |
| p3 | Heisenberg conditioning | Closed-loop cue-preference learning | Yes (needs FicTrac) | ≈8:04 min | ≈18:04 min |
Animated pattern previews are embedded on all four protocol pages.
Also on the rigs
p100_rig_test— a 60-second, controller-only checkout (no fly, no FicTrac): full-field brightness, panel map, a moving bar, and LED levels 2/5/10/20/40%. Use it to confirm a rig works after setup.fictrac_direction_test— open-loop motion then a closed-loop block, to confirm the FicTrac bridge is working.optomotor_led_test— a slow grating with the LED switched on for the last few seconds; the reference example of “fire a command partway through a trial.”
How to read a protocol page
Each page tells you: what the fly sees, what the LED does, how the trials are organized, and roughly how long it takes. The authoritative source is always the protocol’s YAML file (in protocols/shared/), whose header comment describes it in full detail.
TBD: add which genotypes pair with which protocol and the intended order for the course sessions.
Analysis preview
These analysis pages are still being built. The first plots will likely include:
- P0: LED dose-response curves and time-aligned velocity/turning around LED onset.
- P1: optomotor matrices by spatial/temporal frequency and looming matrices by stimulus class/speed/position.
- P2: closed-loop fixation traces, open-loop sweep responses, and side-balanced object-choice preference plots.
- P3: arena-index occupancy, baseline-corrected preference, quadrant dwell times, LED transitions, and walking/immobility QC.
Updated 2026-07-10 01:47 ET.