Experimental pipeline for identifying 9 neuronal subtypes within behaviorally relevant ROIs determined by GCaMP6s imaging.

a, Surgical paradigm. In a TetO-GCaMP6s x CaMKII-tTa mouse, 9 AAVretro viruses are injected into downstream brain regions and GRIN lens implanted into the target region. b, Simultaneous recording of GCaMP6s (top) and behavior (bottom) during a social memory task. Scale bar = 100 µm c, GCaMP6s recordings are processed. CNMF-defined ROIs (top) and ΔF/F traces (bottom) are exported. Scale bar = 100 µm. d, Mice are head fixed and FOV under the GRIN lens imaged using the multiplexed lambda method. e, Transformations are determined using anatomical background images to co-register the two imaging platforms. The transformations are applied to CNMF-defined ROIs. Scale bar = 100 µm. f, Multispectral data are collected for each ROI (top) and an average spectral fingerprint for all ROIs is generated (bottom). Mean ± 1.5 SD. Scale bar = 100 µm. g, A linear unmixing model is applied to determine the fluorophore contribution for each ROI. Scale bar = 100 µm. i, Neural activity is sorted by cell type. Scale bars = 20 ΔF/F (vertical), 20 seconds (horizontal).

GRIN lens induced chromatic aberrations

a, Multicolor image obtained through 1x4 mm silver-doped GRIN lens of the calibration slide highlighting the z-plane chromatic aberration. b, Shift in z-focal plane as a function of excitation laser wavelength. Second-order polynomial R2 = 0.9926, n = 5. c, Orthogonal projection of multicolor image obtained through 1x4mm silver-doped GRIN lens of the calibration slide. Intensity profile (below) of single ring for each excitation channel shows negligible chromatic shift along lateral axes. d, Percent transmission through the GRIN lens as a function of excitation laser wavelength. Sixth order polynomial R2 = 0.9751, n = 5. e, Orthogonal projection of calibration slide imaged through 1x4mm silver-doped GRIN lens overlaid (in cyan) with rectilinear grid lines. Substantial overlap of fluorescent rings from the grid indicates minimal field distortions. f, Excerpt from (e) showing the rings focused in the sagittal plane (z= +20 µm), the circle of least confusion (z= 0 µm), and the tangential focal plane (z= -17.5 µm). g, Curvature of the Petzval field as a function of radial distance from center of the GRIN lens. Astigmatism results in three axially separated focal planes. Second order polynomial sagittal R2 = 0.9845, least confusion R2 = 0.8839, and tangential R2 = 0.7519, n = 3. Scale bars = 100 µm.

Identification of 9 fluorophores through GRIN lenses in vivo

a, In vivo multiplexed spectral imaging paradigm. Schematic of multiplexed spectral imaging (left). Depiction of overlapping fluorophore spectral emissions for each excitation laser wavelength (middle). Depiction of multiplexed spectral images which create a 204-dimensional dataset (right). b, Automated co-registration of miniscope and laser scanning confocal microscope (LSM) images. Top: A calibration slide used to measure scaling between modalities. Bottom: Experimental FOV showing brain vasculature. Miniscope and confocal images of the same FOV and automated co-registration overlay with zoomed-in regions of interest. c, Example calcium-activity ROI derived from miniscope data co-registered and overlaid on confocal LSM image. d, Spectral fingerprint of the example ROI, with the solid blue line showing the example ROI and the dashed line depicting the average spectral profile of the animal. e, Beta multiplier from the example ROI, depicting the deviation from the mean beta value for all ROIs from the same animal. f, Empirically measured spectral profiles from pure fluorophore samples, shown as beta-weighted contributors to ROI fingerprints. Scale bar: 100 µm (a, b), 10 µm (b inset and c).

Distribution of fluorophore-positive functionally defined ROIs

a, Identified fluorophores for injection paradigm A. Viral injection paradigm A consisted of mTagBFP2 into the dPAG, mTurquoise2 into the BLA, T-Sapphire into the Cla, mVenus into the NAc, mOrange2 into the Str, mScarlet into the LC, FusionRed into the VTA, mCyRFP1 into the LHb, and mNeptune2.5 into the c-mPFC (left). Spatial distribution of ROIs and respective fluorophore matches overlaid on anatomical images from the same mouse (right). Distribution of identified fluorophores per mouse (inset). b, Identified fluorophores for injection paradigm B. Viral injection paradigm B consisted of mTagBFP2 into the c-mPFC, mTurquoise2 into the LHyp, T-Sapphire into the Str, mVenus into the VTA, mOrange2 into the LC, mScarlet into the dPAG, FusionRed into the BLA, mCyRFP1 into the NAc, and mNeptune2.5 into the Cla (left). Spatial distribution of ROIs and respective fluorophore assignments overlaid on the anatomical image from the same mouse (right). Distribution of identified fluorophores per mouse (inset). Color and letter codes for fluorophores and injection regions, respectively (far right). c, Percentage of identified ROIs with a fluorophore match. Animal n = 5; ROI n = 1,327. d, Percent of cells identified for each fluorophore. Letter insets on individual data points correspond to injected regions. N = 1,072. One-way ANOVA p = 0.0071. e, Percent of cells identified for each injected region. Number insets on individual data points correspond to injected fluorophore. N = 1,072. Mean ± SEM. One-way ANOVA, p = 0.2599. Scale bars: 100 µm.

Neuronal cell-types vary in behavioral encoding

a, Representative trace (left) and averaged ΔF/F traces (right) of calcium transients time-locked to behavioral annotations. Top traces depict a nucleus accumbens-projecting neuron with annotations denoting social interaction with either a familiar (black) or novel (grey) conspecific. Bottom traces denote a locus coeruleus-projecting neuron with annotations denoting aggressive (black) or investigative (grey) social interactions, regardless of conspecific target. Scale bars = 10 ΔF/F y-axis, 2 s (x-axis, left) and 5 ΔF/F (y-axis), 1 s (x-axis, right). N = 34 behavioral epochs for familiar, 38 epochs for novel, 32 epochs for aggressive, and 53 epochs for investigative interactions. Mean ± SEM. Two-sided t-test: familiar max response vs. novel, p = 0.0011; aggressive vs. investigative, p = 0.049. b, Distribution of the number of different behaviors for which each neuronal cell type statistically modifies its firing rate. c, Schematic of neuronal cell-types and the behavioral categories for which each cell encodes.

AAVretro fluorophore evaluation

Injection coordinates

Comparison of aberrations across GRIN lens types

a, Shift in z-focal plane as a function of excitation laser wavelength in a 1x4 mm silver-doped GRIN lens. Second-order polynomial R2 = 0.9970. b, Shift in z-focal plane as a function of excitation laser wavelength in 0.6x7 mm GRIN lenses doped with either silver or lithium. Second order polynomials: Silver R2 = 0.9983, Lithium R2 = 0.9928. c, Percent transmission through a 1x4 mm silver-doped GRIN lens as a function of excitation laser wavelength. Sixth-order polynomial R2 = 0.9926. d, Percent transmission through 0.6x7 mm GRIN lenses doped with either silver or lithium as a function of excitation laser wavelength. Sixth-order polynomial: silver, R2 = 0.9979; lithium, R2 = 0.9549.

Multiplexed spectral fingerprints for single-fluorophore samples

a, Spectral fingerprints for each fluorescent proteins were obtained by measuring emission intensities using multiplexed spectral imaging in transfected HEK293T cells.

Subject-to-subject variation

a–e, Average spectral fingerprints computed across all ROIs for each experimental mouse. Mean ± 1.5 x SD. f–j, Spatial distribution of ROIs and their corresponding fluorophore assignments overlaid on anatomical images from the same mouse. Scale bars = 100 µm.

Representative examples of fluorophore-identified neurons using Neuroplex

a-r, Each panel shows an ROI that exceeded threshold for a single fluorophore identity assignment. Left: Functional ROIs identified from miniscope recordings during behavior, co-registered and overlaid on corresponding in vivo confocal images. Center: Spectral fingerprint of the ROI (solid line), compared to the animal’s average spectral background (dashed line). Excitation-emission bins are color-coded to excitation laser wavelength. Right inset: Beta multiplier values for all fluorophores from the same ROI, plotted as standard deviations above the animal-specific baseline. Scale bars = 10 µm.

Modeling-based evaluation of fluorophore identification robustness under challenging conditions

a–h, Simulated single-fluorophore datasets were used to assess Neuroplex classification performance under conditions designed to mimic common experimental noise sources. Each panel shows a schematic of the modeling setup (inset), average identification accuracy for single-pass (gray) and dual-pass (black) classification (top), and a breakdown of classification outcomes by fluorophore, including correct identifications, false-negatives, and false-positives (bottom). Each simulation contained 250 ROIs and was repeated 100 times. Modeled perturbations included: increasing GCaMP background within ROIs (a–b), added low-level spectral background from other fluorophores (c–d), decreased signal-to-noise ratio via Gaussian white noise (e–f), and over-representation of a single fluorophore in the population (g–h). i, Comparison of theoretical equal beta contributions (dashed line) to empirically observed beta weights across fluorophores in an example animal, showing deviations that inform the need for adjusted thresholds. j, Example ROI illustrating how second-pass analysis recovers fluorophore assignments missed during initial thresholding. k, Simulated experimental condition using known ROI spectra matched to the actual fluorophore distribution in animal YAS21272R. Background fluorescence, GCaMP co-expression, and white noise were added to approximate in vivo signal characteristics. l, Final classification breakdown for the modeled condition in k, using the dual-pass approach.

Modeling dual fluorophore expressing ROIs

a, Modeled conditions with ROIs containing either single or dual fluorophores: depiction (left), breakdown of dual-pass analysis performance per fluorophore pair (right). b, Modeled experimental conditions with ROIs containing either single or dual fluorophores with added experimental background, GCaMP background, and white noise: depiction (left), breakdown of dual-pass analysis performance (right). Bars reflect percent of ROIs where each fluorophore was correctly identified (colored), misidentified (white), or not identified (grey). ROIs n = 460, replicates n = 100.

Experimental ROIs expressing dual-fluorophores

a, Proportion of functionally defined ROIs classified as expressing one or two fluorophores based on thresholded beta multipliers. b, Frequency of dual fluorophore assignments across the dataset. Left: Heatmap showing co-assignment rates between fluorophore pairs. Right: Total frequency of dual hits per individual fluorophore. c, Frequency of dual-labeled ROIs by brain region. Left: Heatmap showing co-occurrence between projection-defined populations. Right: Total frequency of dual hits per primary brain region. (n = 1,327 ROIs) d, Example ROIs from miniscope imaging co-registered with confocal lambda stacks. ROIs are overlaid on three excitation channels (405, 561, 639 nm). e, Z-scored beta multipliers across all fluorophores for each example ROI, with above-threshold values circled. Dual fluorophores were assigned to two ROIs (28, 98), and only a single fluorophore to ROI 142. f, Spectral fingerprints for each example ROI (solid line) plotted against the average background spectrum for the animal (dashed line). ROI 28 (top) was assigned two spectrally distinct fluorophores (mTagBFP2 + mNeptune2.5); ROI 98 (middle) shows co-assignment of more spectrally overlapping fluorophores (mOrange2 + mNeptune2.5); ROI 142 (bottom) is included as a single-label example with a strong match to mTagBFP2 only. Scale bars = 10 µm.

Neural cell types and behavioral encoding

a, Behavioral paradigm used to identify neurons modulated during social behavior and memory. Data displayed are from the testing phase. b, Schematic of neuronal cell-types and the behavioral categories for which each cell encodes. The top tier identifies the contribution of cells from each animal subject, the second tier and color depict the neural cell-type by projection region, the third identifies the behavior in which the neurons statistically modified their activity, and the fourth determines whether the neurons increased (+) or decreased their activity (-).

Performance of Neuroplex under reduced fluorophore complexity

a, Schematic of experimental design: Four retrograde fluorophores were injected into distinct brain regions of GCaMP6s transgenic mice to label projection-defined pyramidal neurons in the mPFC. mTagBFP2 was injected into the c-mPFC, mVenus into the Str, mOrange2 into the Cla, and mNeptune2.5 into the VTA. b–c, Spatial distribution of identified ROIs overlaid on anatomical reference images (left) and corresponding average spectral fingerprints from each experimental animal (right). Shaded regions represent ±1.5 std from the mean. Scale bars 100 µm. d, Proportion of ROIs classified as single- or dual-labeled based on dual-pass thresholding. Animal n = 2; ROI n = 289. e, Percent of identified ROIs assigned to each individual fluorophore. f, Frequency of dual-fluorophore assignments across the dataset. Left: Heatmap showing pairwise co-occurrence rates between fluorophore (and region) combinations. Right: Total frequency of dual hits per individual fluorophore/region. g–h, Modeled experimental conditions assessing classification accuracy in either single-fluorophore (g) or dual-fluorophore (h) expression contexts. Spectra were simulated with empirical fluorophore distributions, real background, GCaMP contamination, and Gaussian white noise (left). Breakdown of error types (right). ROIs n = 460, replicates n = 100. i, Fluorophore interaction analysis showing classification accuracy when fluorophores were co-expressed. Bars indicate the percentage of ROIs correctly identified (colored), missed (gray: false negatives), or incorrectly labeled (white: false positives).