Skip to main content

BF528 - Applications in Translational Bioinformatics

Project 3: ATACseq Individual Project

This dataset consists of two ATACseq replicates from a human source. Remember that ATACseq datasets typically do not have a control. I will provide some hints since you have not worked with this data type before. This project will be slightly more challenging than the first two.

You will start by performing QC and adapter trimming as you have done for other experiments. I would recommend you align reads using BowTie2 with the -X 2000 flag. After alignment, you will need to first remove any alignments to the mitochondrial chromosome (look at samtools view options).

After this filtering, you will want to shift your reads to account for the bias induced by the “tagmentation” process (there are various tools to accomplish this including deeptools). After this, you will want to perform a quality control analysis by looking at the fragment distribution sizes for your samples using a tool like ATACSeqQC.

You can perform peak calling using MACS3 and their recommended default parameters for ATACseq for each replicate. Similar to ChIPseq, you will need to come up with a bedtool strategy to generate a single set of “reproducible” peaks and then filter any peaks falling into blacklisted regions. After this, you will perform similar analyses to a ChIPseq.

You may find the following publications helpful:

https://www.nature.com/articles/nmeth.2688

https://genomebiology.biomedcentral.com/articles/10.1186/s13059-020-1929-3

Your report must address the following topics/questions in writing:

  1. Briefly remark on the quality of the sequencing reads and the alignment statistics, make sure to specifically mention the following:
    • Are there any concerning aspects of the quality control of your sequencing reads?
    • Are there any concerning aspects of the quality control related to alignment?
    • Based on all of your quality control, will you exclude any samples from further analysis?
  2. After alignment, quickly calculate how many alignments were generated from each sample in total and how many alignments were against the mitochondrial chromosome
    • Report the total number of alignments per sample
    • Report the number of alignments against the mitochondrial genome
  3. After performing peak calling analysis, generating a set of reproducible peaks and filtering peaks from blacklisted regions, please answer the following:
    • How many peaks are present in each of the replicates?
    • How many peaks are present in your set of reproducible peaks? What strategy did you use to determine “reproducible” peaks?
    • How many peaks remain after filtering out peaks overlapping blacklisted regions?
  4. After performing motif analysis and gene enrichment on the peak annotations, please answer the following:
    • Briefly discuss the main results of both of these analyses
    • What can chromatin accessibility let us infer biologically?

Deliverables

  1. Produce a fragment length distribution plot for each of the samples

  2. Produce a table of how many alignments for each sample before and after filtering alignments falling on the mitochondrial chromosome

  3. Create a signal coverage plot centered on the TSS (plotProfile) for the nucleosome-free regions (NFR) and the nucleosome-bound regions (NBR)

    • You may consider fragments (<100bp) to be those from the NFR and the rest as the NBR.
    • You may also wish to separate the fragments into the NFR (<100bp) and the mononucleosome fraction (reads between 180bp and 247bp), which may make this plot more clear.

  4. A table containing the number of peaks called in each replicate, and the number of reproducible peaks

  5. A single BED file containing the reproducible peaks you determined from the experiment.

  6. Perform motif finding on your reproducible peaks

    • Create a single table / figure with the most interesting results

  7. Perform a gene enrichment analysis on the annotated peaks using a well-validated gene enrichment tool

    • Create a single table / figure with the most interesting results

  8. Produce a figure that displays the proportions of regions that appear to have accessible chromatin called as a peak (Promoter, Intergenic, Intron, Exon, TTS, etc.)