scGPT: Fine-tuning on Pre-trained Model for Cell-type Annotation

Cui et al.

scGPT (v0.2.1) is a foundation model for single-cell biology, based on generative pre-trained transformers trained on a vast repository of over 33 million cells. The scGPT model demonstrates the ability to extract valuable biological insights and can be further optimized through transfer learning for various downstream applications, including cell-type annotation, multi-batch integration, genetic perturbation prediction, and gene network inference.

Example use case: Fine-tune a pre-trained model on a new dataset for the cell type annotation task

Technology: Transformers

Limitation:

  • Some of the parameters were kept default. Please see this page for more details.
  • The whole-human model is tested only.
Citation:
scGPT: Towards Building a Foundation Model for Single-Cell Multi-omics Using Generative AI Haotian Cui, Chloe Wang, Hassaan Maan, Kuan Pang, Fengning Luo, Bo Wang bioRxiv 2023.04.30.538439; doi: https://doi.org/10.1101/2023.04.30.538439
Released: Aug-08-2023
v0.1
Single-Cell Bioinformatics
Cell Type Annotation
h5ad
357
27
share
Example Results
Previous Job Parameters
Your previous job parameters will show up here
so you can keep track of your jobs

Upload Single Cell RNA-Seq Reference Data File

No files selected!
OR
Use Demo Data
Your file can be in the following formats:
h5ad
• The h5ad format in scRNA-seq refers to the Hierarchical Data Format 5 (HDF5) Annotated Data format. It is commonly used to store single-cell gene expression data.

Upload Single Cell RNA-Seq Test Data File

No files selected!
OR
Use Demo Data
Your file can be in the following formats:
h5ad
• The h5ad format in scRNA-seq refers to the Hierarchical Data Format 5 (HDF5) Annotated Data format. Test dataset must contain X_UMAP key in the obsm section.

Set Parameters

Select...
Select...
Select...
cell_ranger
human_all
1
1
400
-1
-1
1000
-1
-1
1000
-1
-1
5000