Single Cell RNA-Seq Analysis

scRNA-Seq Analysis

ScRNA-Seq Seurat workflow automation.

Single cell sequencing technologies measure and analyze the molecular signatures in thousands to millions of individual cells. These methods help to characterize the diversity of cell types in simple or complex tissues. They have been used to study the heterogeneity of brain cell types [1], the immunological system cell diversity [2], or the heterogeneity of cell type composition in different cancer types [3].

Nowadays, ultra-high throughput single cell sequencing technologies can sequence millions of cells or nuclei from a single experiment across different tissues, enabling characterizing the molecular signatures for a wide range of diseases at cellular level [4]. A recently introduced approach, cell Hashing allows uniquely labelling of different experimental samples using oligonucleotide-tagged antibodies to detect cell surface proteins [5]. There exist several bioinformatics tools developed to process and analyze the high-dimensional data generated by sequence alignments and feature counting of each dataset. However, the rising accessibility to this technology and the great variability of applications in medicine and biomedical research creates a need for automatic data processing and analysis.

Here we present a scRNA-seq workflow based on Seurat’s scRNAseq workflow for multiplexed samples [6]. This workflow aims to scale the one dataset analysis to an automated and reproducible analysis for a large number of scRNAseq datasets.

Short description of the workflow and its purpose:

In this workflow we analyze Cell Multiplex data. It takes as input single cell FastQ sequencing files and a feature reference, and it uses Cellranger multi pipeline to produce a matrix of UMI counts per {feature|barcode} pair, in MEX format (raw_feature_bc_matrix files) [7]. This matrix is then the input for the Seurat automation workflow.

The workflow first demultiplexes each sample at a time and performs a quality control (QC) step. Samples demultiplexing is then followed by the integration of all samples data and a downstream integrated analysis. Downstream integrated analysis includes: Dimensional reduction and clustering, cell type identification, DEG, trajectory analysis and ontology pathway analysis.

Tools used in this workflow:

ScRNA-Seq techniques

Different techniques are available for generating scRNA data [8][9]:

CEL-seq2/C1
Drop-seq
MARS-seq
SCRB-seq
Smart-seq/C1
Smart-seq2

The difference between these techniques is that:

Some use unique molecular identifiers (UMIs) versus sequencing full length RNA
While others are capturing more cells or more genes per cell

scRNA-Seq scRNASeq Technologies

UMIs are a randomly generated barcode added to sequencing libraries before any PCR amplification steps, which enabling the accurate bioinformatic identification of PCR duplicates.
UMI are a tool for genomic variant detection and by using UMIs instead of actual reads, the output will include less noise and less bias caused by the PCR amplification step.
Refer to [8] and [9].

Bibliography: