scRNA-seq analysis Pipeline starting from 10x data

This repository contains a series of R scripts and Jupyter Notebooks for analyzing single-cell RNA-sequencing data from 10x raw data. The goal of this project was to 1) characterize the differentiation process of iPSCs into GABAergic inhibitory neurons and 2) compare different protocols to determine which produces more mature, forebrain specific inhibitory neruons. This involed comparing a well-characterized protocol involving induced expression of Ascl1 and Dlx2 (Yang et al., 2017) to a combination of Dual SMAD inhibiton and Ascl1/Dlx2 expression. The scripts provided include code for the following processes: Cell filtering and processing, cell clustering, sample integration for batch correction, cell annotation, similarity analysis, pseudotime analysis, trajectory inference, RNA velocity, and identification of key transcription factor regulomes.

This is the basic workflow, with each step in separate scripts:

Cellranger was performed with the following code:

cellranger count --id="AD3" --transcriptome=/mnt/users/linda/references/refdata-gex-GRCh38-and-mm10-2020-A/ \
--fastqs=/mnt/users/linda/Ruiqi_10x_Feb2024/ \
--sample="AD3" --localcores=18

1. Seurat R script contain code for starting from 10x filtered_feature_bc_matrix, creating seurat object, cell and mito filtering, and pre-processing steps.
2. Seurat_analysis contains key features of seurat, such as label transfer methods, integration, cell cycle regression, etc.
3. Integration contains integration methods outside of seurat, such as CSS.
4. Gruffi_ribo_removal contains the code to use the Gruffi package in order to remove stressed cells based on expression of key stress genes.
5. Mrtree was used to determine the optimal number of clusters per sample.
6. The contents of cell annotation contain all the types of annotation packages tested, some of which are cell marker based, and others which compare to gene expression of similar cell transcriptomes.
7. Cell Trajectory and pseudotime analysis to determine the differentiation timeline of our cells.
8. RNA velocity
9. SCENIC to identify key transcription factors involved in the differentiation process and their target genes.
10. Brain Region Similarity Analyses to compare our cells to age-matched fetal cells during development. Created some functions to automate the processing of the reference data