Xen1 pooled AAV SPRA screen in mouse lumbar dorsal horn
Combining Machine Learning and Multiplexed, In Situ Profiling to Engineer Cell Type and Behavioral SpecificityThe full Xen1 pooled AAV library tested 27 candidate regulatory elements and controls, each represented by three unique synthetic barcodes, for a total of 81 designed constructs. After intraspinal delivery to mouse lumbar dorsal horn, barcode RNA was measured with a custom Xenium spatial transcriptomics panel and modeled with RESSCU to estimate cell-type and neuron-subtype specificity.
Processed tables are specific to each experiment. Column names, units, measurements, and table structure are not standardized across the database. Check this experiment’s column definitions and quality-control notes before comparing or combining data.
Perturbation & assay details
Basal / Untreated
Spatial Parallel Reporter Assay (SPRA) used a non-integrating PHP.eB AAV pool carrying approximately 500-bp regulatory elements, each linked to three unique barcodes and a DL5/FAP reporter. Following intraspinal delivery to adult C57BL/6J mouse lumbar spinal cord and at least four weeks of expression, barcode RNA was measured at single-cell resolution with a custom Xenium panel. RESSCU negative-binomial models adjusted for mouse identity, cell area, spatial density of negative controls, and, for neuron-subtype models, hSyn/SV40 crosstalk.
Processed data
50 rows per page. Click a cell to inspect its full value.
Visible columns (31 of 31)
| Row | |||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | |||||||||||||||||||||||||||||||
| 2 | |||||||||||||||||||||||||||||||
| 3 | |||||||||||||||||||||||||||||||
| 4 | |||||||||||||||||||||||||||||||
| 5 | |||||||||||||||||||||||||||||||
| 6 | |||||||||||||||||||||||||||||||
| 7 | |||||||||||||||||||||||||||||||
| 8 | |||||||||||||||||||||||||||||||
| 9 | |||||||||||||||||||||||||||||||
| 10 | |||||||||||||||||||||||||||||||
| 11 | |||||||||||||||||||||||||||||||
| 12 | |||||||||||||||||||||||||||||||
| 13 | |||||||||||||||||||||||||||||||
| 14 | |||||||||||||||||||||||||||||||
| 15 | |||||||||||||||||||||||||||||||
| 16 | |||||||||||||||||||||||||||||||
| 17 | |||||||||||||||||||||||||||||||
| 18 | |||||||||||||||||||||||||||||||
| 19 | |||||||||||||||||||||||||||||||
| 20 | |||||||||||||||||||||||||||||||
| 21 | |||||||||||||||||||||||||||||||
| 22 | |||||||||||||||||||||||||||||||
| 23 | |||||||||||||||||||||||||||||||
| 24 | |||||||||||||||||||||||||||||||
| 25 | |||||||||||||||||||||||||||||||
| 26 | |||||||||||||||||||||||||||||||
| 27 | |||||||||||||||||||||||||||||||
| 28 | |||||||||||||||||||||||||||||||
| 29 | |||||||||||||||||||||||||||||||
| 30 | |||||||||||||||||||||||||||||||
| 31 | |||||||||||||||||||||||||||||||
| 32 | |||||||||||||||||||||||||||||||
| 33 | |||||||||||||||||||||||||||||||
| 34 | |||||||||||||||||||||||||||||||
| 35 | |||||||||||||||||||||||||||||||
| 36 | |||||||||||||||||||||||||||||||
| 37 | |||||||||||||||||||||||||||||||
| 38 | |||||||||||||||||||||||||||||||
| 39 | |||||||||||||||||||||||||||||||
| 40 | |||||||||||||||||||||||||||||||
| 41 | |||||||||||||||||||||||||||||||
| 42 | |||||||||||||||||||||||||||||||
| 43 | |||||||||||||||||||||||||||||||
| 44 | |||||||||||||||||||||||||||||||
| 45 | |||||||||||||||||||||||||||||||
| 46 | |||||||||||||||||||||||||||||||
| 47 | |||||||||||||||||||||||||||||||
| 48 | |||||||||||||||||||||||||||||||
| 49 | |||||||||||||||||||||||||||||||
| 50 |
Filters apply to this table only. The CSV download contains the complete processed table; filtered rows are available through the API.
Column dictionary · 31 definitions
- barcode_model_id
- RESSCU response-variable identifier for the barcode model, preserved exactly from the source workbook.
- element_id
- Library regulatory-element identifier parsed from the response-variable identifier, such as E14.
- element_name
- Human-readable enhancer, promoter, or control name from the supplemental library table.
- element_type
- Interpreted library category: candidate enhancer, promoter control, positive control, or negative control.
- target_group
- Cell type or neuronal target group associated with the element in the supplemental library design.
- design_focus
- Controlled design-focus label; these regulatory elements and controls are region-focused rather than allele-focused.
- barcode_id
- Individual synthetic barcode identifier parsed from the response-variable identifier.
- model_level
- RESSCU model level: cell_type or neuron_subtype.
- term_class
- Classification of the model term as intercept, covariate, or cell_identity.
- effect_term
- Exact RESSCU regression term from the source coefficient table.
- estimate_log_count_effect
- Natural-log coefficient estimate for the term; cell identity coefficients quantify attributable barcode-count effects after covariate adjustment.
- std_error
- Standard error of the coefficient estimate.
- statistic
- Regression test statistic for the coefficient.
- p_value
- Regression p-value for the coefficient.
- conf_low
- Lower confidence bound for the coefficient on the natural-log scale.
- conf_high
- Upper confidence bound for the coefficient on the natural-log scale.
- exp_estimate_fold
- Exponentiated coefficient, calculated during processing as a fold multiplier on the model's count scale; not an RNA/DNA ratio.
- prop_positive
- Fraction of cells positive for the modeled barcode from the corresponding model-fit statistics.
- mcfadden_r2
- McFadden pseudo-R2 for the corresponding cell-type model, when supplied.
- auroc
- Area under the receiver operating characteristic curve for the corresponding model.
- auprc
- Area under the precision-recall curve for the corresponding model.
- mae
- Mean absolute error for the corresponding neuron-subtype model; blank where not supplied for cell-type models.
- null_mae
- Mean absolute error for the corresponding null neuron-subtype model.
- mae_nonzero
- Mean absolute error restricted to nonzero observations for the corresponding neuron-subtype model.
- null_mae_nonzero
- Nonzero-observation mean absolute error for the corresponding null neuron-subtype model.
- uniformity_p
- DHARMa-style residual uniformity diagnostic p-value.
- dispersion_p
- DHARMa-style residual dispersion diagnostic p-value.
- zero_inflation_p
- DHARMa-style zero-inflation diagnostic p-value.
- skewness
- Skewness statistic for the barcode count distribution or model diagnostic, as supplied by the workbook.
- paper_qc_pass
- TRUE for coefficient/model rows retained in the author-supplied RESSCU workbook after the paper's QC; absent designed barcode IDs were not imputed.
- source_sheet
- Name of the source workbook sheet from which the coefficient row was taken.
Quality control
Author-reported QC retained Xenium cells with cell-label transfer confidence >0.5, excluded three small ambiguous neuron clusters, and retained 362,006 cells (63,610 neurons) across 3 Xen1 mice. Five of 81 synthetic barcode probes were excluded after high signal in two virus-free mice; RESSCU fit diagnostics were reported but were not used as per-barcode exclusion thresholds (4% zero-inflation p<0.05, 88% dispersion p<0.05, 48% non-uniformity p<0.05). The processed table retains only the 2,043 coefficient rows for the 75 barcode models supplied in the authors' RESSCU workbook; six designed barcode IDs have no model rows and are not imputed.
Curation notes
This package uses the authors' RESSCU summary outputs rather than a raw single-cell barcode count matrix, which was not included in the open supplementary package. The sequence PDF preserves the 500-bp element and barcode sequences in image-rendered Tables S11-S12. The paper reports five false-positive barcodes removed from the 81-probe panel, while media-4 contains 75 barcode model IDs; six designed IDs are absent from that workbook and no values were invented. Model estimates are natural-log count coefficients; exp_estimate_fold is calculated here and should not be interpreted as a conventional RNA/DNA MPRA activity ratio. The reference genome is null because the assay library combines macaque-, mouse-, human-derived, and synthetic/control elements rather than using one assembly.