Neuropixels Opto: mouse circuit control and optotagging, 2026
Lab or organization
No atlas brief linked yet.
Mouse optogenetics with Neuropixels Opto
The June 2026 published paper uses Neuropixels Opto to record neural activity while delivering spatially addressed red and blue light. It tests laboratory circuit manipulation and identification of genetically defined cell classes. It is not a human assistive-control trial.
Distinct experiments
| Experiment | Evidence scope |
|---|---|
| UCL cortical activation | Acute primary visual cortex recordings in awake, head-fixed mice; 13 recordings in three ChRmine-expressing mice for the reported unit-yield comparison |
| Activation preparation | Four experimental adult mice plus a separate no-opsin control described in Methods; these are not all the yield-analysis cohort |
| Washington cortical circuit manipulation | Three adult mice; nine sessions. Red-sensitive ChrimsonR targeted putative inhibitory neurons with the DLX 2.0 enhancer |
| Allen subcortical optotagging | 261 tagged units across 40 sessions in 26 adult mice (11 male, 15 female), using multiple lines/opsins |
These are separate scopes, not one interchangeable cohort. The paper includes additional expression/validation experiments; it does not justify adding every described preparation into one participant total.
Activation and suppression
Cortical activation used red-sensitive ChRmine and addressed emitters at different depths. The reported yield was 0.23 ± 0.09 units per site (median ± median absolute deviation; 13 recordings in three mice). The comparison 0.22 ± 0.10 is an external Neuropixels 1.0 dataset, 20 recordings in 20 mice in ten labs, selected with the same quality criteria. It is not a matched randomized control cohort.
The circuit experiment activated ChrimsonR-expressing putative inhibitory neurons and recorded suppression of putative excitatory neurons. These labels are not simply proved by waveform shape; the paper discusses expression strategies and circuit responses. Opsin expression is required, and outcomes should not be translated into a clinical electrical-stimulation protocol.
Optotagging
The aggregate 261 tagged units are cell-type identifications over 40 sessions, not the probe’s physical channel count or a chronic tracked-neuron total. A striatal example tags 25 of 39 recorded units; that fraction is one example, not the success rate across all mice. Dual-color experiments target combinations such as D 1/D 2 medium spiny neurons.
Methods require significant response to at least four pulses from one emitter, latency below 8 ms and mean response reliability above 0.3. Units responding to both colors are classified as red-tagged because red-shifted opsins can respond to blue light. Blue-tagged and red-tagged populations are therefore not assigned by wavelength alone.
Recording quality and boundaries
The general cluster-quality criteria are ISI violation ratio below 0.5, amplitude cutoff below 0.1 and presence ratio above 0.8. Different experiments use different Kilosort versions and pipelines; a single sorter is not imposed on the entire paper.
Blue-light leakage, sharp-onset red artifacts and light scattering remain hardware/interpretation limits. No months-long implant-survival curve, human safety demonstration or therapeutic efficacy is established by these acute experiments. Longitudinal reliability should not be borrowed from other Neuropixels variants.