MNase-seq analysis for Choi et al.

bash and R code for MNase-seq analysis performed in Choi, Lyons et al. (2020) (https://www.cell.com/molecular-cell/pdf/S1097-2765(19)30789-0.pdf)

isolate mapped mononucleosomal-sized fragments from sam files
e.g. for 120 to 180bp inclusive:

perl -wnla -e '(abs($F[8])>119 and abs($F[8])<181) and print;' your_sam_file

then run

numap.sh

which is a wrapper of initial numap scripts as provided at
https://github.com/orphancode/NuMap

then run the following to remove non-nuclear genome data:

 perl -wnl -e '/^\d/ and print;' dyad_positions.txt  >  dyad_pos_<uniqueIdentifier>.bg

then put them all together in a single directory and run:

for i in ./*bg; do phasogram_of_sites positions_file=./$i output_file=./$i.phaso max_dist=3000 ; done

then estimate peaks on phasogram output:

for i in ./*phaso; do estimate_peaks dist_file=./$i output_file=./${i%%}.phasogram.smoothed bw=30 field=1 ; done

then import into R and do these analyses with the following scripts:

calculate_NRL.R:

generates a table of NRL estimates from the linear regression fit of the distances between dyad peaks generated in with numap

plot_linear_fit.R:

to visualize the regression line used in calculating the NRL

plot_phasogram.R:

plot the histogram of dyad-to-dyad distances as a smoothed line