Skip to main content

BF528 - Applications in Translational Bioinformatics

Week 1: ChIPseq

Project 2 Directions

Now that we have experience with Nextflow from two prior projects, the directions for this project will be much less detailed. I will describe what you should do and you will be expected to implement it yourself. If you are asked to perform a certain task, you should create working nextflow modules, construct the proper channels in your main.nf, and run your workflow.

Please follow all the conventions we’ve established so far in the course.

These conventions include:

  1. Using isolated containers specific for each tool
  2. Write extensible and generalizble nextflow modules for each task
  3. Encoding reference file paths in the nextflow.config
  4. Encoding sample info and sample file paths in a csv that drives your workflow
  5. Requesting appropriate computational resources per job

Week 1 Overview

For many NGS experiments, the initial steps are largely universal. We perform quality control on the sequencing reads, build an index for the reference genome, and align the reads. However, the source of the data will inform what quality metrics are relevant and the particular choice of tools to accomplish these steps. For RNAseq, it is important to use a splice-aware aligner when aligning against a reference genome since our sequences originated from mRNA. For ChIPseq experiments, our reads originated from DNA sequences and we can use a non-splice aware algorithm to map our reads to the reference genome.

Objectives

  • Assess QC on sequencing reads using FastQC

  • Trim adapters and low-quality reads using Trimmomatic

  • Align trimmed reads to the human reference genome

  • Run samtools flagstat to assess alignment statistics

  • Use MultiQC to aggregate all of the QC metrics

  • Use samtools to sort and index your BAM (alignment) files

Containers for Project 2

FastQC: ghcr.io/bf528/fastqc:latest

multiQC: ghcr.io/bf528/multiqc:latest

bowtie2: ghcr.io/bf528/bowtie2:latest

deeptools: ghcr.io/bf528/deeptools:latest

trimmomatic: ghcr.io/bf528/trimmomatic:latest

samtools: ghcr.io/bf528/samtools:latest

macs3: ghcr.io/bf528/macs3:latest

bedtools: ghcr.io/bf528/bedtools:latest

homer: ghcr.io/bf528/homer:latest

Quality Control, Genome indexing and alignment

Between project 1 and the early labs we did, you should have working modules that perform quality control using FastQC and trimmomatic, build a genome index using bowtie2, and align reads to a genome. We are going to take advantage of the modularity of nextflow by simply copying these previous modules and incorporating them into this workflow.

  1. Copy the subsampled files from /projectnb/bf528/materials/project-2-chipseq/subsampled_files/ to a new directory called subsampled_files/ in your samples/ directory. Encode the names of the file and the paths in a samplesheet at the top level of your directory

  2. Copy the human reference genome and associated GTF to your refs/ directory. Add these paths to your nextflow.config as params.

  3. Update your workflow main.nf to perform quality control using FastQC and trimmomatic, build a bowtie2 index for the human reference genome and align the reads to the reference genome.

Sorting and indexing the alignments

Many subsequent analyses on our BAM files will require them to be both sorted and indexed. Just like for large sequences in FASTA files, sorting and indexing the alignments will allow us to perform much more efficient computational operations on them.

  1. Create a module that will both sort and index your BAM files using Samtools. Do both steps in one single module.

Calculate alignment statistics using samtools flagstat

The samtools flagstat utility will report various statistics regarding the alignment flags found in the BAM.

  1. Create a module that will run samtools flagstat on all of your BAM files

Aggregating QC results with MultiQC

Just like in project 1, we are going to use multiqc to collect the various quality control metrics from our pipeline. Ensure that multiqc collects the outputs from FastQC, Trimmomatic and flagstat.

  1. Make a channel in your workflow main.nf that collects all of the relevant QC outputs needed for multiqc (fastqc zip files, trimmomatic log, and samtools flagstat output)

  2. Copy your previous multiqc module and incorporate it into your workflow to generate a MultiQC report for the listed outputs

Generating bigWig files from our BAM files

Now that we have sorted and indexed our alignments, we are going to generate bigWig files or coverage tracks containing the number of reads per genomic interval or bin for each sample. We will use these coverage tracks for calculating correlation between our samples and visualizing the read coverage in specific regions of interest.

  1. Use the bamCoverage deeptools utility to generate a bigWig file for each of the sample BAM files.

  2. You may use all default parameters. If you wish, you may change the -bs and -p flags as needed.

Week 1 Tasks Summary

  • Create nextflow modules that run the following tools:

    1. FastQC
    2. Trimmomatic
    3. Bowtie2-index

  • Sort and index your bams using a single nextflow module and samtools

  • Calculate alignment statistics using samtools flagstat

  • Aggregate all of the QC output results from the previous tools using MultiQC

  • Generate a nextflow module that uses deeptools to create bigWig representations of your BAM files