Giter VIP home page Giter VIP logo

dlhw's Introduction

scDeepCluster_pytorch

scDeepCluster, a model-based deep embedding clustering for Single Cell RNA-seq data.

alt text

Reference:

Tian, T., Wan, J., Song, Q., & Wei, Z. (2019). Clustering single-cell RNA-seq data with a model-based deep learning approach. Nature Machine Intelligence, 1(4), 191-198. https://www.nature.com/articles/s42256-019-0037-0.

Requirements:

Scanpy -- 1.7

Pytorch -- 1.8

Usage:

python run_scDeepCluster.py --data_file data.h5 --n_clusters 0

Set data_file to the destination to the data (stored in h5 format, with two components X and Y, where X is the cell by gene count matrix, and Y is the true labels. Y is optional), n_clusters to the number of clusters (0 for automatically estimating by the Louvain algorithm on the pretrained latent features).

python run_scDeepClusterBatch.py --data_file data.h5 --n_clusters 0

This is the script for clustering analysis of datasets with batches (stored in h5 format, with three components X, B and Y, where X is the cell by gene count matrix, B is the one-hot encoded batch IDs, and Y is the true labels. Y is optional). n_clusters to the number of clusters (0 for automatically estimating by the Louvain algorithm on the pretrained latent features).

Parameters:

--n_clusters: number of clusters, if setting as 0, it will be estimated by the Louvain alogrithm on the latent features.
--knn: number of nearest neighbors, which is used in the Louvain algorithm, default = 20.
--resolution: resolution in the Louvain algorith, default = 0.8. Larger value will result to more cluster numbers.
--select_genes: number of selected genes for the analysis, default = 0. Recommending to select top 2000 genes, but it depends on different datasets.
--batch_size: batch size, default = 256.
--data_file: file name of data.
--maxiter: max number of iterations in the clustering stage, default = 2000.
--pretrain_epochs: pretraining iterations, default = 300.
--gamma: coefficient of the clustering loss, default = 1.
--sigma: coefficient of the random Gaussian noise, default = 2.5.
--update_interval: number of iteration to update clustering targets, default = 1.
--tol: tolerance to terminate the clustering stage, which is the delta of predicted labels between two consecutive iterations, default = 0.001.
--final_latent_file: file name to output final latent representations of the autoencoder, default = final_latent_file.txt.
--predict_label_file: file name to output clustering labels, default = pred_labels.txt.

Online app

Online app website: https://app.superbio.ai

dlhw

dlhw's People

Contributors

6sy666 avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.