Cell AtlasCSE 496Runs on your device
UNCERTAINTY-AWARE SINGLE-CELL ANNOTATION

Useful labels.
Honest uncertainty.

Find the supported reference, inspect the cell type, and keep the broad label when the fine answer is uncertain.

3reference families
1,320mouse-brain clusters
Localquery inference
Research prototype. Not a diagnostic tool. Confident errors can occur under dataset shift; rejection is not reliable novel-cell detection. Read the limits ↗
Cells processed—
Any label accepted—
Fine label accepted—
No label accepted—
REFERENCE-RELATIVE VIEW

Explore a fitted reference

Loading reference display…

02

Inspect cells

CellReferenceAccepted annotationDepthStatus
Upload a matrix to inspect cell-level predictions. Gray map points are training-reference cells, not uploaded results.
COMPLETED FOLLOW-UP STUDY

The evidence, including the failures.

Seven primary references. 108 neural fits. Twenty directed lung transfers. Reference-label agreement, not independent biological validation.

Bigger is not always better

Primary fine-label accuracy comparison across seven references

Three-seed neural ensembles are compared with PCA+kNN. The small flat ensemble has the best ordinary accuracy on mouse brain and zebrafish; hierarchy is strongest on mouse-brain balanced accuracy.

Keep the supported broad answer

Mouse brain accepted coverage and error at four annotation depths

The selected mouse expert accepts broad class labels for 99.19% of the holdout, at 0.13% error among accepted assignments. Fine cluster coverage is 81.31%, with 8.63% accepted error.

Transfer can be confidently wrong

Accuracy across all 20 directed lung transfers

Aligned gene identifiers are not equivalent expression scales. A controls-trained model accepts every tested 60-day target cell at the fine level despite 50.44% accepted-path error.

Unseen types remain difficult

Fine-label rejection rates for cell types excluded from training

Four of five lung held-out-type tests reject none of the unseen fine types. Good within-reference calibration does not establish dependable novelty detection.

Scope and reproducibility

The follow-up compares small flat, large flat, shared-trunk conditional hierarchical and gene-masked flat networks. It includes six sample/library holdouts, seven held-out-type tests and five leave-one-lung-file-out tests. The earlier 414-specialist hierarchy is a separate historical experiment, not the same architecture.

Methods are selected on tuning balanced accuracy. Test outcomes are not used to replace a selected expert. The follow-up was informed by earlier results; training seeds are not biological replicates, and sample/library groups are not assumed to be donors.

Data source: the completed v2 archive and independent review. The public application deploys selected experts—not every research fit. Detailed diagnostics and model identity are exported with each query.

METHOD, INPUT CONTRACT & LIMITATIONS

Know what the model can—and cannot—say.

1 · Validate the panel

Use UTF-8 CSV with cell_id in the first column and exact gene identifiers in the header. Rows are cells. Or use a JSON object mapping exact genes to numeric values for a single cell. Up to 200 cells, 50,000 genes, 10 million values and 64 MiB per run.

cell_id,gene_A,gene_B,…
cell_1,1.23,0,…
cell_2,0,4.56,…

At least 90% of the relevant selected panel must be present. Duplicate identifiers, ambiguous mapped duplicates, blank/nonfinite values and missing headers are rejected. A measured zero is different from an unmeasured gene. Missing selected genes are imputed using training means. The code does not silently transpose matrices.

2 · Select a reference

Explicit mouse symbol/Ensembl associations harmonize identifiers. A shared-panel, three-member rank-based neural ensemble screens mouse lung versus brain. A separate panel compatibility rule handles zebrafish; it is not a learned species probability. Inputs compatible with both or neither panel are not forced into a family.

The mouse routing confidence is a calibrated model score over those two families—not proof of tissue origin or a probability over all tissues. Reference-support scores describe empirical compatibility with the source; they are not tissue posteriors.

3 · Limit specificity

Child experts are selected using tuning balanced accuracy: a large flat ensemble for pooled lung, and masked-flat ensembles for mouse brain and zebrafish. This is not the historical specialist cascade. The optional PCA+kNN mode uses the same fitted reference bank with its own saved calibration and gates.

A greedy taxonomy-consistent path is accepted only while confidence and support requirements pass. A broad label can remain even when a fine label is withheld. Thresholds are frozen, not adjustable to make a query pass.

4 · Interpret the map carefully

The map displays a fixed sample of training-reference cells. A PCA map uses the first two fitted PCA axes. When UMAP display maps are installed, query locations are inverse-distance barycenters of eight reference neighbors in PCA space—not a new query UMAP and not UMAP.transform.

Reference families have separate coordinate systems. The map is a visualization, not an accuracy test. A cell may appear among reference points while still being rejected by high-dimensional support checks.

Native-scale input only

The initial release expects the expression convention used by each reference. Raw counts, log2 expression, log1p expression and z-scores are not interchangeable. Neither a high overlap score nor a plausible plot verifies normalization equivalence. Do not use this prototype for clinical decisions or claim that it reliably detects arbitrary novel cells.

Unknown-type tests and external transfers demonstrate confident mistakes. A model’s accepted label is a research output, not a biological ground truth.

Privacy, publication, and credits

Inference runs in a dedicated worker on your device. The application contains no query-upload endpoint, trackers or analytics. Hosting still receives ordinary page/model-download requests. Exported model weights, reference embeddings, labels and any installed demo samples are publicly downloadable assets. Do not upload confidential human data to a public or shared computer.

Bence Lukacsy · Stony Brook University · CSE 495/496 honors project · Adviser: Professor Scott Stoller. This independent research prototype is not an endorsed university or clinical service. Data, model and reference-asset redistribution must comply with the source permissions.

Deployment model card ↗ · Release checks ↗ · Source and method notes ↗