科研 Skill

光遗传学实验设计

optogenetics-experiment-design

Designs optogenetic experiments: opsin and wavelength choice, light-delivery and power budgeting, and controls for light and heating artifacts. Use when planning an in vivo manipulation.

发布者 Codar安全审核通过v0.1.0更新于 2026-11-18

SKILL.md

Optogenetics experiment design

Optogenetic experiments fail most often not because the opsin does not work, but because the light delivery, dose, or controls were not planned against the specific opsin's kinetics and action spectrum, or because a light/heating confound was not controlled for. This skill walks through opsin choice, light budgeting, viral/transgenic delivery, and the controls a reviewer will ask for.

When to use this skill

Use it when the user is planning (not already running) an in vivo or ex vivo optogenetic manipulation: choosing an opsin, calculating how much light reaches the target, designing light-only/opsin-only controls, or writing the methods section for an optogenetics experiment.

1. Choose the opsin for the manipulation needed

GoalOpsin classExampleKineticsLight
Fast excitation, single spikesChannelrhodopsinChR2(H134R), ChRmine, CoChRms on/offblue (~470 nm), ChRmine/red-shifted at ~590 nm
Sustained excitation, step-functionStabilized ChRSSFO, CatChseconds-minutes (bistable)blue, can be toggled off with a different wavelength
Fast inhibitionChloride/proton pump or channelHalorhodopsin (NpHR), Archaerhodopsin (Arch), GtACR1/2ms-syellow/green/red (~589 nm NpHR; ~565 nm Arch; ~515 nm GtACR1)
Bistable, low-light inhibitionStep-function inhibitorsSwiChR++bistabletoggled on/off with two wavelengths
Independent two-color controlRed-shifted excitatory + blue-shifted inhibitory, or vice versaChrimsonR + GtACR1variesseparate wavelengths with minimal cross-talk

Match the opsin's action spectrum to the LED/laser available, and check whether the second opsin in a dual-color experiment is also activated at the wavelength used for the first (cross-talk is common, especially with blue light activating red-shifted opsins at high power).

2. Delivery method

  • Viral (AAV): choose serotype for tropism (e.g. AAV5/AAV9 broad, AAV-retro for retrograde labeling from axon terminals, AAV-DJ for broader tropism), promoter for cell-type specificity (CaMKIIa for excitatory neurons, hSyn for pan-neuronal, or a Cre-dependent DIO/FLEX construct combined with a Cre driver line for genetic specificity).
  • Transgenic: Cre/Flp driver lines crossed to a reporter, or Cre-dependent AAV into a driver line, for reproducible, well-characterized cell-type targeting without viral-spread variability.
  • Expression time: allow adequate opsin expression before testing (commonly 2-4 weeks for AAV, longer for weaker promoters); verify expression histologically (native fluorescent tag, e.g. GFP/YFP/mCherry fusion) at the end of the experiment, not just assumed from the injection coordinates.
  • Fiber/LED placement: implant the optical fiber or µLED 0.2-0.5 mm above the targeted structure (light scatters and attenuates with depth; placing the fiber tip directly in the target risks tissue damage from the fiber itself and from local heating).

3. Budget the light dose

Light intensity attenuates strongly with tissue depth due to scattering and absorption. For blue light (473 nm) in brain tissue, intensity falls roughly 50% every ~100-150 µm beyond the fiber tip (Yizhar et al. 2011; Aravanis et al. 2007); use a published tissue-light-transmission calculator (e.g. the Deisseroth lab's online irradiance calculator, or the empirical fits in Yizhar et al. 2011) rather than assuming surface power equals power at the target. Define:

  • fiber/LED output power (mW) and numerical aperture,
  • estimated power density at the target depth (mW/mm²),
  • pulse parameters: wavelength, pulse width, frequency, duty cycle, train duration — chosen to match the opsin's kinetics (a slow opsin does not need sub-ms pulses; a fast one used for following high-frequency spike trains does).

Report all of these in the methods, since "X mW at the fiber tip" alone does not specify the dose delivered to the tissue.

4. Light and heat artifact controls

These are the controls most often missing, and most often requested by reviewers:

  • Opsin-negative / light-on control: same light delivery in animals that received a control virus (fluorophore only, no opsin) or in Cre-negative littermates, to separate the behavioral/physiological effect of light itself (retinal activation via scattered light, heating, animal startle to visible light/sound from the stimulation hardware) from the opsin-mediated effect.
  • Opsin-positive / light-off or sham-light control: opsin-expressing animals with no light, or light below threshold, to control for the virus/surgery/implant itself.
  • Heating: continuous, high-power illumination (especially yellow/red light at powers used for inhibitory opsins, or sustained blue light for SSFO) measurably heats tissue (~0.1-1 °C per 10 mW at typical cortical fiber placements; Stujenske et al. 2015 model and measure this directly) and can itself change neural activity or behavior independent of the opsin. Keep duty cycle and average power as low as the behavioral effect allows, and consider a heat-only control (e.g. a non-light-activated heating element, or data from Stujenske et al. 2015's thermal model for the specific power/duration used).
  • Electrophysiological confirmation: when combining optogenetics with recording, confirm opsin efficacy directly (light-evoked spiking for excitatory opsins, light-evoked suppression for inhibitory opsins) in at least a subset of sessions/animals, rather than inferring efficacy from behavior alone.

5. Dose-response and timing controls

  • Run at least two light intensities (or frequencies) when claiming a graded effect, since many behavioral readouts saturate quickly with opsin manipulations.
  • For closed-loop or state-dependent stimulation (for example, triggering on a detected oscillation or behavior), report the detection latency and verify triggering accuracy against an offline ground truth, analogous to neuro-phase-locked-stim-eval.
  • For experiments with a stimulation electrode or recording during light delivery, check for photoelectric/Becquerel artifacts (light striking the electrode or headstage directly can produce an electrical artifact that mimics a neural response) by including a light-only, no-opsin recording control.

6. Reporting checklist (for methods sections)

Opsin (name, promoter, serotype or driver line, AAV titer and injected volume or transgenic cross), stereotaxic coordinates, expression time, light source (laser/LED, wavelength, fiber core diameter and NA), estimated power density at the target depth and how it was calculated or measured, pulse parameters (width, frequency, train duration, duty cycle), controls used (opsin-negative, light-off, heat), and histological or physiological verification of expression and efficacy.

Prerequisites

  • A defined target region/cell type and a way to verify opsin expression post hoc (histology) or in vivo (optical confirmation, electrophysiology).
  • Calibrated light source: measure output power at the fiber tip with a power meter before and periodically during a study; LED/laser output drifts and fiber coupling losses vary across implants.
  • Institutional animal-use approval (IACUC or equivalent) for the surgical and stimulation procedures — this skill does not grant or substitute for that approval.

Limitations

  • This skill is a research-design aid; it does not model light propagation or heating for a specific animal's anatomy and does not replace a calibrated power measurement or a published tissue-optics model for the exact wavelength and preparation in use.
  • Action-spectrum and kinetic parameters vary by opsin variant and construct; always confirm against the specific published characterization of the exact opsin and promoter combination in use, not the general class.
  • Any live-animal procedure is subject to institutional and regulatory animal-welfare review; this skill does not substitute for IACUC approval or a veterinary safety assessment.

示例提示词

帮我设计一个光遗传学抑制实验,选opsin、算光功率并设置好对照组。

所属插件

神经科学面向神经科学研究:EEG/MEG、fMRI、BIDS、spike sorting、连接组、脑机接口解码评估、神经刺激安全与公开神经数据检索。

安装该插件会一并获得同领域的 45 个 Skill / MCP。

同插件的其他科研 Skill

常见问题

光遗传学实验设计能做什么?

Designs optogenetic experiments: opsin and wavelength choice, light-delivery and power budgeting, and controls for light and heating artifacts. Use when planning an in vivo manipulation。

如何安装光遗传学实验设计?

在科搭 Codar 左侧点击「插件」,在「组件」页签搜索「光遗传学实验设计」并安装;在 Claude Code 或 Codex 中可以发送本页的安装提示词,或下载 Skill 包后解压到 ~/.claude/skills 或 ~/.codex/skills。

光遗传学实验设计需要哪些权限?

该科研 Skill没有声明额外权限;执行具体操作时仍由 Agent 按自身规则请求确认。

光遗传学实验设计经过官方审核吗?

是。SpeedSkills 收录的每个版本都会记录内容摘要与权限声明,当前版本已通过安全审核。