scaLR: a low-resource deep neural network-based platform for single cell analysis and biomarker discovery

What this source contributed

Records in this release that cite this source. A record can cite several sources.

• 2 pipelines
• 1 task
• 1 dataset
• 2 evaluations
• 2 results
• 5 evidence claims

Results

Values as the source printed them. Compare results only under compatible protocols, datasets and metrics.

Claims

Statements this source supports, each with where in the source it was checked.

• Reported Cell-type accuracy for scaLR: Printed value 0.942 (Table 2, scaLR row, Cell type Accuracy column)
• Reported Cell-type accuracy for scVI + scANVI: Printed value 0.939 (Table 2, Svi-tools (scVI & scANVI) row, Cell type Accuracy column)
• scaLR: description: Description Low-resource single-cell annotation platform: it trains on feature subsets to rank gene importance, retrains the final model on the top-K features, and batches samples across CPUs to limit memory. (Reviewed profile (2026-09-16): summary at Abstract (paragraph 1); Results/Performance evaluation of scaLR using all genes as features (paragraph 1); architecture fact at Abstract (paragraph 1); Methods/The scaLR platform/Feature extraction (paragraph 1); description text also read at Abstract)
• scVI + scANVI: description: Description Probabilistic single-cell comparator in the scaLR study: variational autoencoders infer low-dimensional cell representations and the scANVI stage uses annotation labels for cell-type prediction. (Reviewed profile (2026-09-16): summary at Methods/scaLR performance comparison with different pipelines (paragraph 6); Results/Comparison of scaLR extracted top 3,500 features as input for different pipelines (paragraph 1); architecture fact at Methods/scaLR performance comparison with different pipelines (paragraph 6); Methods/scaLR performance comparison with different pipelines (paragraph 3); description text also read at Methods/scaLR performance comparison with different pipelines (scVI-tools description))
• scVI + scANVI: uses model scVI: Uses model scVI (Tables 2 and 3 footnote b: scVI epochs 100 and scANVI epochs 25; Results comparison paragraphs)

Models, benchmarks and other records

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.