Maize single-cell resource

scMaize Single-cell atlas and foundation model

A curated maize single-cell resource for atlas visualization, gene expression query, model access, and data download.

Integrated UMAP
Cells385,675
Cell types46
Projects20
Tissues7
20+Projects
385,675Cells
66Samples
7Tissues
3Platforms
Resources

Database modules

D

Data

Dataset records, tissue sources, cell numbers, platforms, publications, and download links.

A

Atlas

Integrated UMAP, tissue distribution, cell type annotation, and marker gene expression.

M

Models

scMaize model versions, pretrained weights, and evaluation results.

O

Apps

Annotation, embedding extraction, expression imputation, and gene similarity tools for query datasets.

Download

Expression matrices, metadata, integrated embeddings, model weights, and example files.

?

Help

Quick start, atlas guide, model guide, citation, FAQ, and contact information.

Data Resources

Integrated maize single-cell and single-nucleus transcriptomic datasets.

20+Projects
385,675Cells
66Samples
7Tissues
3Platforms

Tissue coverage

RootLeafEarSeedlingEndospermTasselShoot

Dataset summary

SourcePublic datasets
Data typesc/snRNA-seq
OutputMatrix / Metadata / Embedding
Loading datasets...
ProjectTissueSamplesPlatformPublicationResource
Loading...

scMaizeAtlas

Interactive maize single-cell atlas with tissue, cell type, and gene expression layers.

Interactive visualizationIntegrated UMAP
Loading atlas data...

scMaize Models

Foundation models for maize single-cell representation and prediction.

GitHub README

Loading model performance...

Online Apps

Submit maize single-cell query data to scMaize tools, track asynchronous jobs, and retrieve result files from a private status link.

Run scMaize analyses without local setup

The Apps portal accepts query files, places jobs into a shared queue, and returns a status link for each submission. Results can be downloaded after the job finishes.

Use the dedicated Apps portal for uploads, job tracking, and result download. The main database page remains focused on browsing atlas resources.

Launch SCmaize Apps
A

Cell Type Annotation

Predict tissue and cell type labels for query maize single-cell or single-nucleus expression data using the scMaize reference atlas.

Input: query .h5adOutput: annotation CSV with confidence scores
Open annotation tool
E

Cell Embedding

Extract 384-dimensional scMaize cell representations for downstream visualization, clustering, retrieval, or model-based comparison.

Input: query .h5adOutput: NumPy embedding matrix
Open embedding tool
I

Expression Imputation

Estimate missing or sparse gene expression values with gene-neighbor relationships learned by the scMaize model.

Input: query .h5adOutput: imputed .h5ad matrix
Open imputation tool
S

Gene Similarity

Search functionally similar maize genes from the scMaize gene embedding space using one or more query gene symbols.

Input: gene listOutput: ranked similarity table
Open similarity tool

Download

Data files, model weights, embeddings, and example datasets.

Loading resources Reading the database file index

Help

Guide, citation, and contact information.

Quick start

Use Data to browse projects, samples, tissues, platforms, publications, and linked source records. Open Atlas for UMAP exploration, Models for pretrained model resources, and Download for metadata, H5AD, model weights, reports, RDS files, and expression matrices.

Data guide

Search by project, sample, tissue, PubMed, or platform. Open a project row to view sample-level metadata, source links, and available files. Dataset filters can be combined for narrower browsing.

Atlas guide

The Atlas page shows a lightweight UMAP preview generated from the integrated scMaizeAtlas. Color points by cell type, major cell type, tissue, project, or variety, and use filters to focus on selected tissues, cell types, or projects.

Models guide

The Models page summarizes scMaizeExp and scMaizeGO resources. Performance values are synchronized daily from the project GitHub README.

Download guide

Use Download for global resources such as metadata, atlas H5AD, and model weights. For sample-specific files, open the Data detail modal and download raw matrices, reports, or filtered RDS files from each sample.

Citation

Please cite scMaize when using atlas data, embeddings, model weights, or derived results. Until a formal citation is listed, link to the project repository: yjthu/scMaize.

Contact us

For data, model, or website questions, use the protected email link below or open the project repository.

CitationGitHubContactFAQ