Celltypist Cell Annotation

CellTypist Cell Type Annotation

Overview

CellTypist is an automated cell type classifier for single-cell RNA-seq data built on logistic regression models trained on curated reference atlases. Given a normalized AnnData object, it predicts cell type labels at the single-cell level and optionally applies majority voting within user-defined clusters to produce consensus, biologically coherent annotations. The tool ships with 45+ ready-to-use models spanning pan-immune, organ-specific, and developmental contexts, and supports training custom models from labeled data.

When to Use

• Annotating PBMC, whole-blood, lymph node, or other immune cell datasets using a single standardized reference model
• Generating a first-pass cell type annotation before manual curation with canonical marker genes
• Annotating cluster-level cell types in published or in-house datasets using majority voting to smooth noisy per-cell predictions
• Comparing annotation results across multiple tissue-specific models to determine the most biologically relevant reference
• Training a custom CellTypist model from a labeled reference dataset for a tissue or species not covered by pre-built models
• Quantifying annotation confidence to flag low-certainty cells (confidence score < 0.5) for manual review or exclusion
• Use scVI/scANVI (scvi-tools-single-cell) instead when you need probabilistic label transfer with batch correction and uncertainty quantification via a variational autoencoder
• Use popV (popv-cell-annotation) instead when you want ensemble consensus from 10+ methods including deep learning and KNN-based approaches

Prerequisites

• Python packages: celltypist>=1.6, scanpy>=1.9, anndata
• Data requirements: AnnData with normalized, log1p-transformed counts in adata.X (10,000 UMIs per cell target sum). Raw counts must be normalized before calling CellTypist
• Environment: Python 3.8+; 8 GB RAM sufficient for most datasets; internet access required for model downloads (first run only)

pip install celltypist "scanpy[leiden]" anndata

Quick Start

Minimal pipeline — annotate a preprocessed AnnData with the pan-immune model:

import celltypist
import scanpy as sc

# Load a preprocessed AnnData (normalized + log1p, Leiden clusters already in adata.obs)
adata = sc.read_h5ad("preprocessed_pbmc.h5ad")

# Run annotation with majority voting across Leiden clusters
predictions = celltypist.annotate(
    adata,
    model="Immune_All_Low.pkl",
    majority_voting=True,
)
adata = predictions.to_adata()

print(adata.obs[["predicted_labels", "majority_voting", "conf_score"]].head(10))
# predicted_labels  majority_voting  conf_score
# CD4+ T cells      CD4+ T cells     0.92
# ...

Workflow

Step 1: Installation and Model Setup

Install CellTypist and download pre-trained models. Models are cached locally after the first download.

pip install celltypist "scanpy[leiden]" anndata

import celltypist
from celltypist import models

# Download all available models (only needed once; ~2 GB total)
models.download_models(force_update=False)

# List available models with metadata
models_df = models.models_description()
print(models_df[["model", "description", "n_celltypes", "n_cells"]].to_string())
# Output (excerpt):
#   model                          description                                 n_celltypes  n_cells
#   Immune_All_Low.pkl             Pan-immune low-hierarchy (98 cell types)   98           324,320
#   Immune_All_High.pkl            Pan-immune high-hierarchy (30 cell types)  30           324,320
#   Human_Lung_Atlas.pkl           Lung cell types from Human Lung Atlas       61           584,944

Step 2: Data Preparation

CellTypist requires normalized, log1p-transformed counts in adata.X. Run normalization before annotation. Raw counts must be stored separately.

import scanpy as sc

# Load raw count matrix
adata = sc.read_h5ad("raw_counts.h5ad")
# Alternatively from 10X:
# adata = sc.read_10x_mtx("filtered_feature_bc_matrix/")
# adata.var_names_make_unique()

# Store raw counts before normalization
adata.layers["counts"] = adata.X.copy()

# Normalize to 10,000 UMIs per cell and log1p-transform
sc.pp.normalize_total(adata, target_sum=1e4)
sc.pp.log1p(adata)

print(f"Prepared: {adata.n_obs} cells x {adata.n_vars} genes")
print(f"adata.X mean: {adata.X.mean():.3f}  (expected ~0.5–2.0 after log1p normalization)")

Step 3: Model Selection

Choose the model that best matches your tissue type and desired annotation resolution.

from celltypist import models

# Show full model table with filtering
models_df = models.models_description()

# Filter to human immune models
immune_models = models_df[models_df["description"].str.contains("immune|Immune", case=False)]
print(immune_models[["model", "description", "n_celltypes"]].to_string())

# Load a specific model to inspect its cell type labels
model = models.Model.load("Immune_All_Low.pkl")
print(f"Model cell types ({len(model.cell_types)}):")
print(model.cell_types[:20])  # first 20 labels

Available models (key selection guide):

| Model | Cell Types | Best For |

|-------|-----------|---------|

| Immune_All_Low.pkl | 98 | Pan-immune with fine subtypes (e.g., MAIT, Tfh, cDC1) |

| Immune_All_High.pkl | 30 | Pan-immune major lineages (T, B, NK, monocyte, DC) |

| Human_Lung_Atlas.pkl | 61 | Lung: alveolar, stromal, immune, endothelial |

| Pan_Fetal_Human.pkl | 139 | Fetal human multi-organ development |

| Developing_Human_Brain.pkl | 51 | Brain development: progenitors, neurons, glia |

| Human_Colorectal_Cancer.pkl | 62 | Colorectal cancer cells + tumor microenvironment |

Step 4: Automated Annotation

Run celltypist.annotate() with majority_voting=True for cluster-level consensus labels alongside per-cell predictions.

import celltypist
import scanpy as sc

# Ensure Leiden clusters exist for majority voting
# If not already computed:
sc.pp.highly_variable_genes(adata, n_top_genes=2000)
sc.pp.pca(adata)
sc.pp.neighbors(adata, n_pcs=30)
sc.tl.leiden(adata, resolution=0.5, key_added="leiden")

# Run CellTypist annotation
predictions = celltypist.annotate(
    adata,
    model="Immune_All_Low.pkl",
    majority_voting=True,          # cluster-level consensus
    over_clustering="leiden",      # clustering key for majority voting
    p_thres=0.5,                   # cells below threshold → "Unassigned"
    mode="best match",             # assign the single highest-probability label
)

# Inspect prediction object
print(type(predictions))  # celltypist.classifier.AnnotationResult
print(predictions.predicted_labels.head())
print(predictions.probability_matrix.shape)  # (n_cells, n_cell_types)

Step 5: Results Integration

Transfer predictions back to the AnnData object and review confidence scores.

# Merge predictions into adata.obs
adata = predictions.to_adata()

# Key result columns:
# adata.obs["predicted_labels"]  — per-cell best-match label
# adata.obs["majority_voting"]   — cluster-level consensus label
# adata.obs["conf_score"]        — probability of the predicted label (0–1)

print(adata.obs[["predicted_labels", "majority_voting", "conf_score"]].head(10))
print(f"\nCell type distribution (majority voting):")
print(adata.obs["majority_voting"].value_counts().head(15))

# Flag low-confidence cells
low_conf = adata.obs["conf_score"] < 0.5
print(f"\nLow-confidence cells (conf_score < 0.5): {low_conf.sum()} ({low_conf.mean():.1%})")
adata.obs["high_conf"] = ~low_conf

Step 6: Visualization and Validation

Plot predictions on UMAP, validate with canonical marker genes, and confirm annotation quality.

import scanpy as sc
import matplotlib.pyplot as plt

# Compute UMAP if not already done
if "X_umap" not in adata.obsm:
    sc.tl.umap(adata)

# UMAP colored by annotation results
fig, axes = plt.subplots(1, 3, figsize=(21, 6))
sc.pl.umap(adata, color="majority_voting", legend_loc="on data",
           legend_fontsize=7, title="Majority Voting", ax=axes[0], show=False)
sc.pl.umap(adata, color="predicted_labels", legend_loc="right margin",
           legend_fontsize=7, title="Per-Cell Prediction", ax=axes[1], show=False)
sc.pl.umap(adata, color="conf_score", cmap="RdYlGn",
           title="Confidence Score", ax=axes[2], show=False)
plt.tight_layout()
plt.savefig("celltypist_annotation.png", dpi=150, bbox_inches="tight")
plt.show()
print("Saved celltypist_annotation.png")

# Validate with canonical immune markers
marker_genes = {
    "CD4+ T": ["CD3D", "CD4", "IL7R"],
    "CD8+ T": ["CD3D", "CD8A", "GZMK"],
    "B cells": ["MS4A1", "CD79A"],
    "NK cells": ["GNLY", "NKG7"],
    "CD14 Mono": ["CD14", "LYZ"],
}
sc.pl.dotplot(adata, var_names=marker_genes, groupby="majority_voting",
              use_raw=False, standard_scale="var",
              save="_celltypist_markers.png")

Key Parameters

| Parameter | Default | Range / Options | Effect |

|-----------|---------|-----------------|--------|

| model | — | Any .pkl filename or path | Selects the reference atlas for annotation; must match tissue/species |

| majority_voting | False | True, False | When True, smooths per-cell labels to cluster consensus; requires a clustering key in over_clustering |

| over_clustering | None | Any adata.obs key, "leiden", "louvain" | Clustering column used for majority voting; auto-detected if common keys present |

| p_thres | 0.5 | 0.0–1.0 | Minimum probability to assign a label; cells below threshold are labeled "Unassigned" |

| mode | "best match" | "best match", "prob match" | "best match": top label regardless of threshold; "prob match": applies p_thres |

| min_prop | 0.0 | 0.0–1.0 | For majority voting: minimum fraction of cluster cells with the consensus label; rare labels may be suppressed |

Key Concepts

Pre-Trained Model Architecture

Each CellTypist model is a one-vs-rest logistic regression classifier trained on a curated cell atlas. Key properties:

• Input: 33,694 genes (or fewer if the dataset has a smaller gene space — unshared genes are zero-filled)
• Output: per-cell probability vector over all cell type classes; highest probability is the predicted label
• Confidence score: the probability assigned to the winning class (0–1); high values (>0.7) indicate reliable predictions
• Species/version specificity: models are trained on specific atlases; using a human model on mouse data will produce spurious results

Majority Voting

Majority voting applies a two-stage correction after per-cell prediction:

• Each cell receives a per-cell label from the logistic regression output
• Within each cluster (e.g., Leiden cluster), the most frequent per-cell label becomes the cluster's consensus majority_voting label
• Cells whose per-cell label disagrees with the cluster majority are re-labeled to the cluster consensus unless min_prop is set

Majority voting is recommended when individual cells have noisy expression but the cluster is biologically coherent. Disable it when cells within a cluster are biologically heterogeneous (e.g., transitional states).

Gene Space Alignment

CellTypist automatically intersects the model's training genes with the input AnnData's gene names. Genes present in the model but absent from the query are zero-filled. Annotations degrade if fewer than ~60% of model genes are present — check with model.cell_types and adata.var_names.

Common Recipes

Recipe: Train a Custom Model

When to use: your tissue or species is not covered by an existing model, and you have a labeled reference dataset.

import celltypist
import scanpy as sc

# Load labeled reference AnnData (must be normalized + log1p)
ref = sc.read_h5ad("labeled_reference.h5ad")
# ref.obs["cell_type"] must contain string cell type labels

# Train custom model
new_model = celltypist.train(
    ref,
    labels="cell_type",       # obs column with training labels
    n_jobs=4,                  # parallel workers
    max_iter=200,              # logistic regression iterations
    use_SGD=False,             # use full L-BFGS-B solver (recommended for <100k cells)
    top_genes=500,             # number of most informative genes per class
)

# Save for reuse
new_model.write("custom_tissue_model.pkl")
print(f"Trained model: {len(new_model.cell_types)} cell types")

# Apply to query
predictions = celltypist.annotate(query_adata, model="custom_tissue_model.pkl",
                                  majority_voting=True)

Recipe: Multi-Model Comparison

When to use: uncertain which model best matches your dataset; run multiple models and compare agreement.

import celltypist
import pandas as pd

model_names = ["Immune_All_High.pkl", "Immune_All_Low.pkl", "Human_Lung_Atlas.pkl"]
results = {}

for model_name in model_names:
    preds = celltypist.annotate(adata, model=model_name, majority_voting=True)
    adata_tmp = preds.to_adata()
    key = model_name.replace(".pkl", "")
    results[key] = adata_tmp.obs["majority_voting"].values

comparison = pd.DataFrame(results, index=adata.obs_names)
print("Agreement between Immune_All_High and Immune_All_Low:")
agreement = (comparison["Immune_All_High"] == comparison["Immune_All_Low"]).mean()
print(f"  {agreement:.1%} of cells agree")
print(comparison.head(10))

Recipe: Export Annotations for Downstream Analysis

When to use: saving annotated data with all prediction metadata for downstream differential expression or trajectory analysis.

import scanpy as sc
import pandas as pd

# Save full annotated AnnData
adata.write_h5ad("annotated_celltypist.h5ad", compression="gzip")
print(f"Saved annotated_celltypist.h5ad  ({adata.n_obs} cells)")

# Export cell type table
cell_table = adata.obs[[
    "predicted_labels", "majority_voting", "conf_score", "leiden"
]].copy()
cell_table.to_csv("celltypist_annotations.csv")

# Cell type proportions per sample
if "sample" in adata.obs.columns:
    props = (adata.obs.groupby(["sample", "majority_voting"])
             .size().unstack(fill_value=0))
    props_norm = props.div(props.sum(axis=1), axis=0)
    props_norm.to_csv("celltypist_proportions.csv")
    print(f"Cell type proportions saved (shape: {props_norm.shape})")

Expected Outputs

| Output | Description |

|--------|-------------|

| adata.obs["predicted_labels"] | Per-cell best-match label from logistic regression |

| adata.obs["majority_voting"] | Cluster-consensus label (when majority_voting=True) |

| adata.obs["conf_score"] | Probability of the predicted label (0–1); >0.5 = confident |

| adata.obsm["X_umap"] | UMAP embedding (if computed in preprocessing step) |

| celltypist_annotation.png | UMAP panels: majority voting label, per-cell label, confidence scores |

| celltypist_annotations.csv | Per-cell annotation table with predicted labels and confidence |

Troubleshooting

| Problem | Cause | Solution |

|---------|-------|----------|

| ValueError: adata.X does not appear to be log1p normalized | Raw counts passed directly | Run sc.pp.normalize_total(adata, target_sum=1e4) then sc.pp.log1p(adata) before calling celltypist.annotate() |

| Many cells labeled "Unassigned" | p_thres too high or model species mismatch | Lower p_thres to 0.3; verify model matches species and tissue; check conf_score distribution |

| KeyError for over_clustering key | Clustering column name not found in adata.obs | Run sc.tl.leiden(adata, key_added="leiden") first, or set over_clustering="leiden" explicitly |

| Implausible labels (e.g., immune labels on neurons) | Wrong model selected for tissue | Choose a tissue-specific model (e.g., Developing_Human_Brain.pkl for brain data); list options with models.models_description() |

| MemoryError on large datasets (>500k cells) | Full probability matrix held in RAM | Subsample to 200k cells for annotation, then transfer labels via KNN; or use mode="best match" to skip storing full probability matrix |

| Low overall conf_score (<0.4 median) | Dataset is poorly represented by the reference model | Train a custom model from a matched reference or use popv-cell-annotation for ensemble voting |

| Model not found error on download | Network issue or wrong model name | Run models.download_models(force_update=True); verify name with models.models_description()["model"].tolist() |

Related Skills

• scanpy-scrna-seq — preprocessing pipeline (QC, normalization, clustering) that produces the AnnData input for CellTypist
• popv-cell-annotation — ensemble annotation using 10+ methods; use when you want consensus across methods rather than a single model
• scvi-tools-single-cell — scANVI for semi-supervised label transfer with deep generative models and probabilistic uncertainty
• harmony-batch-correction — batch correction to apply before annotation when integrating multiple samples

References

• CellTypist documentation — official API reference, model descriptions, and tutorials
• GitHub: Teichlab/celltypist — source code and issue tracker
• Dominguez Conde et al., Science 2022 — "Cross-tissue immune cell analysis reveals tissue-specific features in humans", original CellTypist paper
• CellTypist model portal — interactive model browser with cell type hierarchies and training dataset details