Nextflow Pipeline for 10x MultiOme processing with measurement of heteroplasmy using mgatk

Author: Malwina Prater

Description

Nextflow pipeline for processing 10x MultiOme datasets with CellRanger-arc and measuring mitochondrial heteroplasmy from ATAC-seq.

Analyze 10x multiome (RNA + ATAC) in one pipeline.
Starting point are fastq files

Software requirements

Software	Version
nextflow	v.21.10.6
multiqc	v.1.14
mgatk	v.0.6.6
qualimap	v.2.2.1
fastqc	v.0.11.8
fastq_screen	v.0.14.1
cellranger-arc	v.2.0.1
bwa	v.0.7.15
java	min v.1.8

Cell Ranger arc was downloaded from 10x (https://support.10xgenomics.com/single-cell-multiome-atac-gex/software/downloads/latest).

You will need to change paths to the software and genomes in the nextflow config file.

Processing

Set up a working directory, where you store fastq files, sample table and nextflow config:

Input files

Sample table written in .csv

with formatting like:

Sample_ID,Sample_Project,Sample_Species,Sample_Lib,Sample_Pair
Sample2,ProjectID,mouse,rna,2
Sample2,ProjectID,mouse,atac,2
Sample1,ProjectID,mouse,rna,1
Sample1,ProjectID,mouse,atac,1

For NUMTs masked genome, it has to be specified in sample table as below:

Sample_ID,Sample_Project,Sample_Species,Sample_Lib,Sample_Pair
Sample2,ProjectID,mouse_NUMTs_masked,rna,2
Sample2,ProjectID,mouse_NUMTs_masked,atac,2
Sample1,ProjectID,mouse_NUMTs_masked,rna,1
Sample1,ProjectID,mouse_NUMTs_masked,atac,1

Organise directory with folder fastq and subfolders atac and rna (fastq.gz files can be soft linked):

fastq --|---rna ---|--Sample2_S1_L001_I1_001.fastq.gz
        |          |--Sample2_S1_L001_I2_001.fastq.gz
        |          |--Sample2_S1_L001_R1_001.fastq.gz
        |          |--Sample2_S1_L001_R2_001.fastq.gz
        |
        |---atac---|--Sample2_S1_L001_I1_001.fastq.gz
                   |--Sample2_S1_L001_R1_001.fastq.gz
                   |--Sample2_S1_L001_R2_001.fastq.gz
                   |--Sample2_S1_L001_R3_001.fastq.gz

Create nextflow config file using template nextflow.config, and modify directories and paths to genomes and software.

Prepare genome.

This is already done for Mouse genome in directories (also unmasked genome for Human too):

/suffolk/WorkGenomicsE/mn367/Genomes/CellRanger_Genomes/refdata-cellranger-arc-mm10-2020-A-2.0.0
/suffolk/WorkGenomicsE/mn367/Genomes/CellRanger_Genomes/refdata-cellranger-arc-mm10-2020-A-2.0.0_MT_masked
/suffolk/WorkGenomicsE/mn367/Genomes/CellRanger_Genomes/refdata-cellranger-arc-GRCh38-2020-A-2.0.0

For NUMTs masking of the genome mm10.full.backlist.bed was downloaded from https://github.com/caleblareau/mgatk/wiki/Increasing-coverage-from-10x-processing.

Mask genome using script: mask_NUMTs_genome.sh.

Once the masked fasta is ready, use script CR_mkref.sh to make new CellRanger-arc genome. The config file was required to proceed with making new genome.

How to run the multiome nextflow pipeline

Run nextflow on MBU cluster:

change directory to project dir, where fastq files are stored (or linked).

Before running the pipeline, change software paths, e.g. MultiQC, FastQ-screen, qualimap, CellRanger-arc, mgatk etc.


#module load nextflow/22

module load anaconda
module load qualimap

screen -S <session_name>

NXF_VER=21.10.6 nextflow run nf_pipeline_multiome.nf -config nextflow.config -resume

nmalwinka/Nextflow_Pipeline_MultiOme_Processing