PCLDA: An interpretable cell annotation tool for single-cell RNA-sequencing data based on simple statistical methods
Bibliographic record
Abstract
Single-cell RNA sequencing (scRNA-seq) enables high-resolution analysis of cellular heterogeneity, yet accurate and consistent cell-type annotation remains a crucial challenge. Numerous automated tools exist, but their complex modeling assumptions can hinder reliability across varied datasets and protocols. We propose PCLDA, a pipeline composed of three modules: t-test-based gene screening, principal component analysis (PCA) and linear discriminant analysis (LDA), all built on simple statistical methods. An ablation study shows that each module in PCLDA contributes significantly to performance and robustness, with two novel enhancements in the second module yielding substantial gains. Despite these additions, the model retains its original assumptions, computational efficiency, and interpretability. Benchmarking against nine state-of-the-art methods across 22 public scRNA-seq datasets and 35 distinct evaluation scenarios, PCLDA consistently achieves top-tier accuracy under both intra-dataset (cross-validation) and inter-dataset (cross-platform) conditions. Notably, when reference and query data are generated via different protocols, PCLDA remains stable and often outperforms more complex machine-learning approaches. Furthermore, PCLDA offers strong interpretability, attributed to the linear nature of its PCA and LDA modules. The final decision boundaries are linear combinations of the original gene expression values, directly reflecting the contribution of each gene to the classification. Top-weighted genes identified by PCLDA better capture biologically meaningful signals in enrichment analyses than those selected via marginal screening alone, offering deeper functional insights into cell-type specificity. In conclusion, our work underscores the utility of carefully enhanced simple statistics methods for single-cell annotation. PCLDA's simplicity, interpretability, and consistently high performance make it a practical, reliable alternative to more complex annotation pipelines. Code is available on GitHub:https://github.com/kellen8hao/PCLDA.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.010 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.004 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| Insufficient payload (model declined to judge) | 0.008 | 0.006 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".