Appendix 1—table 1. Standardization.
a, To facilitate reproducibility we standardized multiple aspects of the experiment. Some variables were kept strictly the same across mice, while others were kept within a range or simply recorded (see ‘Standardized’ column). b-c, The behavior training protocol was also standardized. Several task parameters adaptively changed within or across sessions contingent on various performance criteria being met, including number of trials completed, amount of water received and proportion of correct responses.
| A | ||||
|---|---|---|---|---|
| Category | Variable | Standardized | Standard | Recorded |
| Animal | Weight | Within a range | 18–30 g at headbar implant | Per session |
| Age | Within a range | 10–12 weeks at headbar implant | Per session | |
| Strain | Exactly | C57BL/6J | Once | |
| Sex | No | Both | Once | |
| Provider | Two options | Charles River (EU) Jax (US) | Once | |
| Training | Handling | One protocol | Protocol 2 | No |
| Hardware | Exactly | Protocol 3 | No | |
| Software | Exactly | Protocol 3 | Per session | |
| Fecal count | N/A | N/A | Per session | |
| Time of day | No | As constant as possible | Per session | |
| Housing | Enrichment | Minimum requirement | At least nesting and house | Once |
| Food | Within a range | Protein: 18–20%, Fat: 5–6.2% | Once | |
| Light cycle | Two options | 12 Hr inverted or non-inverted | Once | |
| Weekend water | Two options | Citric acid water or measured water | Per session | |
| Co housing status | No | Co-housing preferred, separate problem mice | Per change | |
| Surgery | Aseptic protocols | One protocol | Protocol 1 | No |
| Tools/Consumables | Required parts | Protocol 1 | No | |
| B | ||||
| Adaptive parameter | Initial value | |||
| Contrast set | [100, 50] | |||
| Reward volume | 3 μL | |||
| Wheel gain | 8 deg/mm | |||
| C | ||||
| Criterion | Outcome | |||
| >200 trials completed in previous session | Wheel gain decreased 4 deg/mm | |||
| >80% correct on each contrast | Contrast set = [100, 50, 25] | |||
| >80% correct on each contrast after above | Contrast set = [100, 50, 25, 12.5] | |||
| 200 trials after above | Contrast set = [100, 50, 25, 12.5, 6.25] | |||
| 200 trials after above | Contrast set = [100, 50, 25, 12.5, 6.25, 0] | |||
| 200 trials after above | Contrast set = [100, 25, 12.5, 6.25, 0] | |||
| 200 trials completed in previous session and reward volume > 1.5 μL | Decrease reward by 0.1 μL | |||
| Animal weight/25 > reward vol/1000 and reward volume < 3 μL | Next session increase reward by 0.1 μL | |||
| D | ||||
| Proficiency level | Outcome | |||
| ‘Basic task proficiency’ - Trained 1a/1b For each of the last three sessions: >200/400 trials completed, and >80%/90% correct on 100% contrast and all contrasts introduced and For the last three sessions combined: psychometric absolute bias < 16/10 and psychometric threshold < 19/20 and psychometric lapse rates < 0.2/0.1 For 1b only: median reaction time at 0% contrast < 2 s |
Training in the basic task achieved: mouse is ready to proceed to training in the full task. In some mice, we continued training in the basic task to obtain even higher performance. (Guideline was to train mice for up to 40 days at Level 1, and drop mice from the study if they did not reach proficiency in this period). |
|||
| ‘Full task proficiency’ For each of the last three sessions: >400 trials completed, and >90% correct on 100% contrast and For the last three sessions combined: all four lapse rates (left and right, 20:80 and 80:20 blocks)<0.1 and bias[80:20] - bias[20:80]>5 and median RT on 0% contrast < 2 s |
Training in the full task achieved: mouse is ready for neural recordings. | |||