ATAC-seq pipeline with snakemake

Created 18-Feb-2021

Before the pipeline, here are considerations for experiment design:

• two or more biological replicates
• each replicate has 25 million non-duplicate, non-mitochondrial aligned reads for single-end sequencing and 50 million for paired-ended sequencing
• typically, no need for “input”
• use as few PCR cycles as possible when constructing the library
• paired-end sequencing is preferred

Basic steps of ATACseq

from here

Reads to Peaks

• trim reads with cutadapt
• Alignment with bowtie2
• sort with samtools
• filter out mito genes with samtools
• mark pcr duplicates with Picard's MarkDuplicates
• remove duplicates with samtools
• remove multi-mapped reads with samtools
• Remove reads unmapped, mate unmapped, not primary alignment, reads failing platform, duplicates (-F 1804) with samtools
• Retain properly paired reads -f 2 with samtools
• (optional) check correlations between BAM files with deeptools multiBamSummary
• (optional) merge bams with samtools
• (optional) analyses that require single-base resolution e.g TF motif footprinting) shift reads with bedtools/awk
• (optional) filter for nucleosome-free regions
• Call peaks with MACS2 (use -B parameter for creating browser tracks)
• Annotate peaks with ChIPseeker or Homer

QC

• Count fragments in bam files. Should have at least 25 million non-duplicate, non-mito fragments
• Check alignment rate, should be > 95%
• Measure library complexity with Non-Redundant Fraction (NRF) and PCR Bottlenecking Coefficients 1 and 2, or PBC1 and PBC2.
• Calculate IDR values (Irreproducible Discovery Rate). Rescue and self consistency rations should be < 2
• Calculate TSS enrichment and FRiP score
• Plot fragment size distribution

Params file

Necessary information

• filebase
• SE or PE sequencing (default to PE)
• reference
• comparisons
• replicates

Other notes

• For now, consider running both HOMER and MACS2
• Make heatmap with deeptools
• Different ways of dealing with replicates (combine bam files and call peaks, or merge peaks?)
• diff peaks with Homer or edgeR (which can run in homer)
• setup directory structure to have project/data/reads... and project/analysis/reference/peaks...
• generate README.md
• be sure to load modules with version number