CVRCseq

CVRCseq is a unified Snakemake workflow collection for common NGS analyses on Slurm-based HPC systems (developed for NYU UltraViolet).

Available Workflows

RNA-seq

• RNAseq_PE: paired-end, fastqc -> fastp -> STAR -> featureCounts
• RNAseq_SE: single-end, fastqc -> fastp -> STAR -> featureCounts
• RNAseq_PE_HISAT2_stringtie: paired-end, fastqc -> fastp -> HISAT2 -> StringTie
• RNAseq_PE_HISAT2_stringtie_nvltrx: paired-end, fastqc -> fastp -> HISAT2 -> StringTie -> novel transcript workflow
• RNAseqTE_PE: paired-end, fastqc -> fastp -> STAR -> TEcount

Small RNA-seq

• sRNAseq_SE: single-end, fastqc -> umi-tools -> STAR -> featureCounts

DNA Binding / Enrichment

• ChIPseq_PE: paired-end, fastqc -> fastp -> bowtie2 -> MACS2
• CUT-RUN_PE: paired-end, fastqc -> fastp -> bowtie2 -> MACS2
• ATACseq_PE: paired-end, fastqc -> fastp -> bowtie2 -> MACS2

Repository Structure

• workflow/Snakefile: top-level workflow entry point; loads one rules file based on workflow in config.
• workflow/rules/*.smk: per-workflow rule definitions.
• workflow/scripts/snakemake_init.sh: main launcher script.
• workflow/scripts/cat_rename.py: optional preprocessing step for lane concatenation and FASTQ renaming.
• config/config.yaml: global and workflow-specific parameters.
• config/samples_info.tab: sample metadata table.
• config/profile/config.yaml: Snakemake profile and Slurm defaults.
• workflow/envs/CVRCseq.yml: conda environment definition.

Configuration

Sample Metadata (config/samples_info.tab)

Expected columns include:

1. FASTQ file names (R1/R2)
2. User-friendly sample name
3. Condition
4. Replicate
5. Antibody/control label (required for ChIP-seq and CUT-RUN)
6. Final sample ID (used for renamed FASTQ output)
7. Optional additional metadata

Notes:

• cat_rename.py concatenates multi-lane FASTQs and renames files from this table.
• For ChIP-seq and CUT-RUN pairs, keep sample name/condition/replicate consistent between IP and control rows.

Main Config (config/config.yaml)

Common keys:

• sample_file: path to sample table (default config/samples_info.tab)
• workflow: active workflow name (set automatically by snakemake_init.sh)
• genome: index path (STAR, HISAT2, or bowtie2 depending on workflow)
• GTF: annotation file path

Workflow-specific keys:

• CUT-RUN_PE:
  • spike_genome
  • chromosome_lengths
  • effective_genome_size
• ChIPseq_PE, ATACseq_PE:
  • effective_genome_size
• RNAseq_PE_HISAT2_stringtie, RNAseq_PE_HISAT2_stringtie_nvltrx:
  • prepDE_length
  • stringtie_strandedness (example: "--rf")
• RNAseqTE_PE:
  • TE_GTF
  • TE_strandedness (example: "reverse")
• RNAseq_PE, RNAseq_SE, sRNAseq_SE:
  • featurecounts_strandedness (0, 1, or 2)

Running the Pipeline

1) Clone

git clone https://github.com/mgildea87/CVRCseq.git
cd CVRCseq

2) Prepare inputs

• Update config/samples_info.tab.
• Update config/config.yaml for your references and workflow settings.

3) Launch

bash workflow/scripts/snakemake_init.sh -d /path/to/fastq -w RNAseq_PE

Options:

• -h: help
• -d: FASTQ directory (required)
• -w: workflow name (required)
• -s: extra Snakemake args (quote multiple flags, for example -s "--dryrun --quiet")
• -c: skip cat_rename.py
• -i: override Singularity image path

If needed, unlock a stale Snakemake directory:

snakemake --unlock --profile config/profile

This requires loading the container or conda evironment where snakemake is installed

Execution Mode (Container vs Host)

Default behavior:

• Uses Singularity image at /gpfs/data/cvrcbioinfolab/shared_conda_envs/CVRCseq.sif if available.
• Falls back to host conda environment (/gpfs/data/cvrcbioinfolab/shared_conda_envs/CVRCseq) if the image is absent and -i is not provided.

Pull the image manually:

module load singularity/3.11.5
singularity pull --dir /gpfs/data/cvrcbioinfolab/shared_conda_envs/ docker://mgildea87/cvrcsseq:latest

For additional container details, see container/README.md.

Running on a Compute Node

Launching from a compute node is recommended. Update workflow/scripts/launch_sbatch.sh and submit:

sbatch workflow/scripts/launch_sbatch.sh

Tool Links

• Snakemake
• STAR
• FastQC
• fastp
• subread/featureCounts
• HISAT2
• StringTie
• TEtranscripts/TEcount
• UMI-tools
• bowtie2
• MACS2