Skip to content

zhanglab-wbgcas/CIForm

Repository files navigation

CIForm

image

Introduction

CIForm, a Transformer-based model, can annotate cell types.

Instructions and examples are provided in the following tutorials.

Requirement

Python 3.9.12
PyTorch >= 1.5.0
numpy
pandas
scipy
sklearn
Scanpy
random

Input file

reference dataset.

cell type label of reference dataset.

query dataset.

Output file

After training the CIForm model, the model will be save at: "log/CIForm.tar".
The model prediction is saved in the log/y_predicts.npy.

Usage

import CIForm as CI
pred_result = CI.ciForm(s, referece_datapaths, Train_names, Testdata_path,Testdata_name)

in which

  • s=The length of sub-vector,
  • referece_datapaths=The path of annotated scRNA-seq datasets
  • Train_names=The name of annotated scRNA-seq datasets
  • Testdata_path=The path of query scRNA-seq datasets
  • Testdata_name=The name of query scRNA-seq datasets

It is recommended that The label file be in the same directory as the corresponding data set and be named Labels.csv The label file should be a n rows * 1 column vector. For example, image

Tutorial

processsed Data

The scRNA-seq datasets pre-processing code

Paper Link

CIForm as a Transformer-based model for cell-type annotation of large-scale single-cell RNA-seq data https://academic.oup.com/bib/advance-article/doi/10.1093/bib/bbad195/7169137

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published