zhangxf-ccnu / scmoo_experiments_codes

Introduction

README file for codes to reproduce the three downstream analyses (such as differential expression analysis, cell clustering analysis and pseudotime analysis) in the paper "Imputing dropouts for single-cell RNA sequencing based on multi-objective optimization". The method scMOO developed in the paper can be found at: https://github.com/Zhangxf-ccnu/scMOO.
To reproduce the masking and down-sampling experiments, please refer to Chen and Zhou (2018): Vpaper2018.

Contents of this archive

This archive contains:

(1) Datasets: subdirectory that contains four preprocessed datasets: bulk data and single-cell data of H1_DEC (such as H1_DEC_bulk and H1_DEC_sc), PBMC_CL and Deng datasets, which can be used to reproduce the three downstream analyses (such as differential expression analysis, cell clustering analysis and pseudotime analysis) respectively.
Note that H1_DEC and PBMC_CL datasets have been preprocessed using Seurat v3.2 to contain 2,000 highly variable genes, and Deng dataset has been preprocessed by filtering out genes expressed in less than 10% of cells.

(2) DE_analysis: subdirectory that contains three R codes to reproduce the differential expression analysis.

Step1_Prepare_datasets.R: After downloading the bulk data and single-cell data of Cell Type (GSE75748) from GEO website, selecting 6 pairs of cell subpopulations including DEC, then using Seurat v3.2 to select 2,000 highly variable genes of both the bulk data and single-cell data.

Step2_edgeR.R: Using edgeR to identify differential expression genes (DEGs) between pairs of cell subpopulations.

Step3_Compute_AUCscores_SpearmanCorrelation.R: Computing AUC scores and Spearman correlation coefficient.

(3) Cell_clustering: subdirectory that contains two R codes to reproduce the cell clustering analysis.

Step1_Select_HVGs.R: Using Seurat v3.2 to select 2,000 highly variable genes of the single-cell data.

Step2_SC3_Seurat.R: Using SC3 and Seurat to carry out cell clustering analysis. And ARI and NMI are used to evaluate the consistency between the results of SC3 or Seurat and the reference labels of cells.

(4) Cell_trajectory_inference: subdirectory that contains three R codes to reproduce the pseudotime analysis.

Step1_Preprocess.R: Obtaining the preporcessed Deng dataset with settings percent=0.1 and preprocess.only=TRUE.

Step2_Monocle2.R: Using Monocle2 to carry out pseudotime analysis. The function returns POS and Kendall's rank correlation scores.

Step3_Trajectory_inference.R: Running Monocle2 on preprocessed dataset with corresponding setting cellLabels.

Tutorial

A tutorial with example of cell clustering analysis at the same time illustrating the usage of scMOO is available at: scMOO-tutorial

Contact

Please do not hesitate to contact Miss Ke Jin [email protected] or Dr. Xiao-Fei Zhang [email protected] to seek any clarifications regarding any contents or operation of the archive.