usage

The pipeline needs to use NextFlow, Singularity/Docker/Conda (for container). Details are given in corresponding pages.

• It is tested locally on Ubuntu server 24.04.2 LTS.

• Tested on HPC (Dardel from PDC, KTH)

Computational Requirements

Software/s Requirements

• Nextflow (>=21.10.3)

• Docker or Singularity (for containerized execution)

• Java (>=8)

Hardware Requirements

• RAM: 12 GB - 200 GB

• CPU: min. 12 cores

• Storage: ~2TB for 24 paired-end samples

Run NextFlow

nextflow run JD2112/TwistMethylFlow \
    -profile singularity \
    --sample_sheet Sample_sheet_twist.csv \
    --genome_fasta /data/reference_genome/hg38/hg38.fa \ 
    --run_both_methods \
    --gtf_file /data/Homo_sapiens.GRCh38.104.gtf \
    --refseq_file /data/hg38_RefSeq.bed.gz \
    --outdir Results/TwistMethylFlow_both 

??? info "additional options" Run with pre-build reference genome index --bismark_index /data/reference_genome/hg38/

**Run only EdgeR**
`--diff_meth_method edger`

**Run only MethylKit**
`--diff_meth_method methylkit`

??? warning "need annotation files" Remember to add annotation files for different differential methylation analysis

`--gtf_file /data/Homo_sapiens.GRCh38.104.gtf` for **MethylKit** 

`--refseq_file /data/hg38_RefSeq.bed.gz` for **EdgeR**

??? note "Help" nextflow run main.nf --help --outdir .

Parameters configuration

User can change the conf/params.config or use the -- flags directly on nextflow command line.

options	Description
--sample_sheet	Path to the sample sheet CSV file (required)
--bismark_index	Path to the Bismark index directory (required unless --genome or --aligned_bams is provided)
--genome	Path to the reference genome FASTA file (required if --bismark_index not provided)
--aligned_bams	Path to aligned BAM files (use this to start from aligned BAM files instead of FASTQ files)
--refseq_file	Path to RefSeq file for annotation (reuired to run both or methylkit)
--gtf_file	Path to GTF file for annotation (reuired to run both or edger)
--outdir	Output directory (default: ./results)
--diff_meth_method	Differential methylation method to use: 'edger' or 'methylkit' (default: edger)
--run_both_methods	Run both edgeR and methylkit for differential methylation analysis (default: false)
--skip_diff_meth	Skip differential methylation analysis (default: false)
--coverage_threshold	Minimum read coverage to consider a CpG site (default: 10)
--logfc_cutoff	Differential methylation cut-off for Volcano or MA plot (default: 1.5)
--pvalue_cutoff	Differential methylation P-value cut-off for Volcano or MA plot (default: 0.05)
--hyper_color	Hypermethylation color for Volcano or MA plot (default: red)
--hypo_cutoff	Hypomethylation color for Volcano or MA plot (default: blue)
--nonsig_color	Non-significant color for Volcano or MA plot (default: black)
--compare_str	Comparison string for differential analysis (e.g. "Group1-Group2")
--top_n_genes	Number of top differentially methylated genes to report for GOplot (default: 100)
--help	Show this help message and exit
-r	Run with --tag version from GitHub (e.g, -r 1.0.5)

??? info "MethylKit specific parameters" --assembly - user needs to provide the genome assembly version. Default: hg38

`--mc_cores` - if required to run on multiple cores. Default: `1`

`--diff` - Differential methylation cutoff value. Default: `0.5`

`--qvalue` - qvalue respect to differential methylation value to identify significant CpGs. Default: `0.05`

??? note "Other parameters configuration" 1. conf/resource.config - for resource settings. 2. conf/base.config - for base settings. 3. nextflow.config - for nextflow run with default setting. 4. dag.config - for DAG configuration settings.

Input File

Sample sheet (CSV format) with sample information

Sample_sheet.csv:

sample_id	group	read1	read2
12A	VD	FASTQ/12A_S9_L001_R1_001.fastq.gz	FASTQ/12A_S9_L001_R2_001.fastq.gz
13A	CS	FASTQ/13A_S11_L001_R1_001.fastq.gz	FASTQ/13A_S11_L001_R2_001.fastq.gz
1A	CS	FASTQ/1A_S1_L001_R1_001.fastq.gz	FASTQ/1A_S1_L001_R2_001.fastq.gz
20A	VD	FASTQ/20A_S13_L001_R1_001.fastq.gz	FASTQ/20A_S13_L001_R2_001.fastq.gz
21A	VD	FASTQ/21A_S15_L001_R1_001.fastq.gz	FASTQ/21A_S15_L001_R2_001.fastq.gz
22A	VD	FASTQ/22A_S17_L001_R1_001.fastq.gz	FASTQ/22A_S17_L001_R2_001.fastq.gz
23A	VD	FASTQ/23A_S19_L001_R1_001.fastq.gz	FASTQ/23A_S19_L001_R2_001.fastq.gz
25A	CS	FASTQ/25A_S21_L001_R1_001.fastq.gz	FASTQ/25A_S21_L001_R2_001.fastq.gz
26A	VD	FASTQ/26A_S23_L001_R1_001.fastq.gz	FASTQ/26A_S23_L001_R2_001.fastq.gz
2A	CS	FASTQ/2A_S3_L001_R1_001.fastq.gz	FASTQ/2A_S3_L001_R2_001.fastq.gz
3A	VD	FASTQ/3A_S5_L001_R1_001.fastq.gz	FASTQ/3A_S5_L001_R2_001.fastq.gz
5A	CS	FASTQ/5A_S7_L001_R1_001.fastq.gz	FASTQ/5A_S7_L001_R2_001.fastq.gz

Results

Here is how the result folder looks like -

.
├── annotate_results
├── bismark_align
├── bismark_deduplicate
├── bismark_genome_preparation
├── bismark_methylation_extractor
├── bismark_report
├── edger_analysis
├── fastqc
├── go_analysis
├── multiqc
├── pipeline_info
├── post_processing
├── qualimap
├── samtools_index
├── samtools_sort
└── trim_galore

??? info "bismark alignment with bowtie2" | Option | Functionality | | ---------------------- | --------------------------------------------------------------------------- | | -q | Quiet mode: suppresses detailed output. | | --score-min L,0,-0.2 | Sets a linear minimum score for valid alignments (moderate stringency). | | --ignore-quals | Ignores base quality scores during alignment. | | --no-mixed | Ensures both ends of paired reads align properly; no single-end alignments. | | --no-discordant | Prevents discordant alignments; enforces proper orientation and distance. | | --dovetail | Allows overlapping or extended alignments in paired-end reads. | | --maxins 500 | Sets the maximum allowed distance between paired-end reads to 500 bases. |

??? tip "SLURM script to run the sample data on HPC cluster" Here is an example SLURM script to run the pipeline on a HPC cluster with Singularity:

```bash
#!/bin/bash    
#SBATCH --job-name=TMF_test
#SBATCH --partition=standard
#SBATCH --nodes=2
#SBATCH --mem=96G
#SBATCH --time=3-00:00:00  # 3 days (D-HH:MM:SS)
#SBATCH --cpus-per-task=16
#SBATCH --output=tmf_test%j.out
#SBATCH --error=tmf_test%j.err
#SBATCH --mail-user=[e-mail]  
#SBATCH --mail-type=BEGIN,END,FAIL 

set -e

# Load modules
module load singularity-4.1.1
module load nextflow-25.04.0

# Set Nextflow and Singularity directories
export NXF_HOME=<Your Nextflow home directory, e.g., /home/username/.nextflow >
export NXF_WORK=<Your Nextflow home directory, e.g., /home/username/nextflow_work>
export SINGULARITY_CACHEDIR=<Your Singularity cache directory, e.g., /home/username/singularity_cache>

# Ensure directories exist
mkdir -p $NXF_WORK $SINGULARITY_CACHEDIR


# Run Nextflow
nextflow run JD2112/TwistMethylFlow \
    -profile singularity \
    --sample_sheet samplesheet.csv \
    --genome_fasta <ABSOLUTE\ PATH\ TO\>/reference_genome/hg19.fa \
    --run_both_methods \
    --refseq_file hg19_RefSeq.bed.gz \
    --gtf_file Homo_sapiens.GRCh37.75.formatted.gtf \
    --outdir TMF_250519 \
    -with-dag
```