Biological replicates are merged directly after filtering. Peak visualization is performed by a custom Gaussian smoother, which creates a Gaussian distribution around the midpoint of each read. Locations with more overlapping reads are thus represented by larger peaks.

The visualization output tracks are bedGraph files, which are converted into bigWig format using bedGraphToBigWig.

The final step of the pipeline is peak calling using SEACR version 1.3. The bedGraph files processed from both experimental and control experiments are needed to call peaks. In particular:

SEACR_1.3.sh <exp_bedgraph> <ctl_bedgraph> <norm> <stringency>

norm indicates the normalization experimental and control data and can be either norm (normalization occurs) or non (normalization does not occur). The default for this pipeline is norm.
stringency specifies the threshold for peak calling. The default for this pipeline is relaxed, which compares the knee and peak of the curve, rather than stringent, which uses the peak of the curve only.

SEACR output is a bed file containing peaks. Each peak is detailed by its chromosome, start and end coordinates, total signal, and the location and value of the maximum signal within its coordinates. Peak calls can be viewed (alongside bigwig tracks) using HiGlass.

SEACR is designed specifically for CUT&RUN experiments. Learn more about SEACR from the SEACR GitHub page.

PageCUT&RUN Processing Pipeline

released View JSON

October 5th, 2021 at 8:07pm

Details

Page
CUT&RUN Processing Pipeline

released
View JSON