SIC-ChIP User Manual

SIC-ChIP works on UNIX and MAC systems

Requirements:

• R (version >3.1.2)
• R packages rtracklayer (downloadable from ​Bioconductor), parallel, BH and Rcpp (downloadable from ​CRAN)
• Boost C++ libraries (downloadable from ​http://www.boost.org/)

Usage:

• Uncompress SIC-ChIP.tar
• Open a new terminal and change directory to the folder where your input files coverage_test.bw and peaks_test.bed are located. The test file coverage_test.bw can be downloaded from ​HERE
• Run SIC-ChIP with the following command:

/path/to/SIC-ChIP.r --bw=coverage_test.bw --bed=peaks_test.bed --out=results [options]

or alternatively:

Rscript /path/to/SIC-ChIP.r --bw=coverage_test.bw --bed=peaks_test.bed --out=results [options]

Mandatory options:

--bw

BigWig file with the coverage function

--bed

BED file with the coordinates of the peaks

--out

The name of the results folder/files

Options:

--help

print short help message and exit

--N

the distance between spline knots for computing the number of local maxima (default --N=20)

--toll

the minimum distance allowed between local maxima (default --toll=50)

IMPORTANT: the parameters N and toll should be carefully chosen depending on the peaks dataset (for example, they should be lower than the default when some peaks have low width)

Output:

A folder named SIC-ChIP_results containing

• The RData file with a RangedData of all the peaks locations and counts, i.e. the coverage function for each peak (results_peaks.RData)
• The RData file with a dataframe of the five shape indices for all the peaks (results_peaks_index.RData)
• The RData file with a list 'K' of k-means results for clusters number k=1,...10 and a vector 'within' with the total within clusters sum of squares plot for cluster number k=1,...,10 (results_kmeans.RData)
• The scatterplot of the five shape indices for all the peaks (results_scatterplot.pdf)
• The total within clusters sum of squares plot (results_tot_within_ss.pdf)

User Support

Please ask a question to marziaangela.cremona[at]polimi.it or mac78[at]psu.edu if needed.