Introduction

immunespace/scintegrator is a bioinformatics pipeline to analyze single-cell RNA-seq (scRNA-seq) data using the Scanpy toolkit. It automates key steps of scRNA-seq analysis, such as data preprocessing, integration, clustering, annotation and visualization, enabling efficient and reproducible workflows.

Pipeline summary

Each of these steps in scintegration pipeline is customizable, allowing researchers to adjust parameters based on the specific characteristics of their dataset and research questions.

• Scanpy_QC - Detail the preprocessing steps implemented in Scanpy, including doublet removal, rigorous Quality Control (QC), and the generation of comprehensive plot reports.

  • Cell Filtering: Remove cells that do not meet certain quality metrics, such as a minimum number of genes expressed or an excessive percentage of mitochondrial gene expression.
  • Gene Filtering: Exclude genes that are not detected in a sufficient number of cells, which helps in focusing the analysis on biologically relevant data.
  • Doublet Detection: Use Scrublet to identify and remove potential doublets from the data, ensuring that each cell analyzed represents a single biological cell.
  • Data Summarization: Generate plots of quality metrics to visually assess the quality of the data and the effectiveness of preprocessing steps.

• Scanpy_Clustering - Scanpy clustering workflow, including data normalization, log transformation, removal of TR and IG genes, identification of highly variable genes, PCA analysis, cell clustering, data integrtion and cell type annotation.

  • Removal of TR and IG Genes: Exclude T-cell receptor (TR) and immunoglobulin (IG) genes which can skew analysis due to their high variability and cell-type specific expression.
  • Data Normalization: Normalize data to make the gene expression profiles more comparable across cells.
  • Log Transformation: Apply log transformation to normalize the distribution of gene expression data.
  • Highly Variable Genes (HVGs) Identification: Select genes with high variability across cells which are informative for clustering.
  • Principal Component Analysis (PCA): Reduce dimensionality of the dataset to capture the most significant gene expression changes.
  • Clustering: Group cells based on similarities in their gene expression profiles using algorithms like Leiden or Louvain.
  • Data Integration: Integrate data from different batches or experiments to correct for variations not related to biological differences using Harmony.
  • Annotation: Annotate identified clusters to known cell types based on marker gene expression.

Quickstart

[!NOTE] If you are new to Nextflow and nf-core, please refer to this page on how to set-up Nextflow and a container engine such as docker or apptainer. Make sure to test your setup with -profile test before running the workflow on actual data.

First, prepare a samplesheet in CSV format with your single-cell expression data in h5 format. The file should look like this:

samplesheet.csv:

csv sample_id,path,study_id sample1,sample1.h5,test sample2,sample2.h5,test

• sample: Sample identifiers.
• path: path to h5 file of scRNA-seq data.
• study_id: Cumstom study name.

Now, you can run the pipeline using:

bash nextflow run immunespace/scintegrator \ -profile docker \ --input samplesheet.csv \ --outdir <OUTDIR>

For further usage details, please refer to the Usage documentation. To understand the pipeline outputs, please refer to the Output documentation. A full list of the pipeline parameters is available in the next section.

Parameters

[!WARNING] Please provide pipeline parameters via the CLI or Nextflow -params-file option. Custom config files including those provided by the -c Nextflow option can be used to provide any configuration except for parameters see docs.

Scanpy_QC - --scanpyspecies [string] Species of the data. [default: human] - --scanpymingenes [integer] Minimum number of genes expressed required for a cell to pass filtering. [default: 300] - --scanpymincells [integer] Minimum number of cells expressed required for a gene to pass filtering. [default: 5] - --scanpypctmt [integer] Maximun percentage of counts in mitochondrial genes required for a cell to pass filtering. [default: 20] - --scanpytotalcounts [integer] Minimum number of the total counts required for a cell to pass filtering. [default: 200] - --qcnb [string] Path to the qc notebook. [default: assets/pipeline_QC.ipynb]

Scanpy_clustering - --expecteddoubletrate [number] The estimated doublet rate for the data. [default: 0.06] - --ensembligtrgenes [string] Immunogloblin (IG) and T cell receptor (TR) genes list. [default: assets/triggenesensemblv111human.csv] - --ensemblspecies [string] Species of the data. [default: human] - --clusteringnneighbors [integer] Size of local neighborhood used for manifold approximation. [default: 10] - --clusteringnpcs [integer] Number of PCs. [default: 40] - --clusteringresolution [number] The resolution of the clustering. [default: 0.5] - --hvgminmean [number] Cutoff for the min means. [default: 0.0125] - --hvgmaxmean [integer] Cutoff for the max means. [default: 3] - --hvgmindisp [number] Cutoff for the normalized dispersions. [default: 0.5] - --clusternb [string] Path to the clustering notebook. [default: assets/pipelinecluster.ipynb]

Input/output options - --input [string] Path to comma-separated file containing information about the samples in the experiment. - --outdir [string] The output directory where the results will be saved. You have to use absolute paths to storage on Cloud infrastructure. - --email [string] Email address for completion summary. - --multiqc_title [string] MultiQC report title. Printed as page header, used for filename if not otherwise specified.

Generic options - --multiqcmethodsdescription [string] Custom MultiQC yaml file containing HTML including a methods description.

Credits

immunespace/scintegrator was originally written by Jian Xing, Gisela Gabernet.

Contributions and Support

If you would like to contribute to this pipeline, please see the contributing guidelines.

Citations

An extensive list of references for the tools used by the pipeline can be found in the CITATIONS.md file.

This pipeline uses code and infrastructure developed and maintained by the nf-core community, reused here under the MIT license.

The nf-core framework for community-curated bioinformatics pipelines.

Philip Ewels, Alexander Peltzer, Sven Fillinger, Harshil Patel, Johannes Alneberg, Andreas Wilm, Maxime Ulysse Garcia, Paolo Di Tommaso & Sven Nahnsen.

Nat Biotechnol. 2020 Feb 13. doi: 10.1038/s41587-020-0439-x.