Mega_dSMF

Analysis pipeline of cytosine methylation single-molecule footprinting (dSMF) experiments with Nanopore sequencing using ONT's Megalodon.

Installation

1. Log on the GPU node (IZBDelhi). This is important to make use of the faster scratch folder. Make sure that you have Conda installed with conda --version
2. git clone https://github.com/AAnnan/Mega_dSMF
3. Review carefully varSettings.sh and update the variables according to your data.
4. Launch the install of the Mega_dSMF conda environment with bash 00_Setup.sh, you will be asked to input Guppy's latest version. You can check it here: https://community.nanoporetech.com/downloads/guppy/release_notes

Usage

Launch the scripts GPU node (IZBDelhi) with sbatch in numerical order. Alternatively, you can launch the whole pipeline in one go with bash MegaWrapper.sh.

One or Two-pass Demultiplexing [deepbinner is deprecated, two_pass will be set to "no" by default]

To demultiplex your raw reads with 01a_Demux_DeepBinner.sh, you must have used one of these sequencing/barcoding kits: EXP-NBD103, EXP-NBD104 or very similar. If you have used a different kit, skip 01a_Demux_DeepBinner.sh and demultiplex only with Guppy by setting the variable two_pass to no in varSettings.sh.

Output

Outputs will be in a folder named output in the Mega_dSMF folder.

Running all scripts (in order) will output:

1. Demultiplexed, basecalled, multifast5s.
2. All Megalodon outputs listed here and in Megalodon_Output_Notes.txt and selected in varSettings.sh
3. 2 BigWig files per barcode of 1-Methylated Fraction along the genome (raw and averaged with a 10-bp rolling window).
4. 3 plots per barcode of the probability distribution of methylation per motif (dSMF, CpG & GpC).

Remarks

• 2 sites (GC and CG) will be explored for 5mC through 3 non-overlapping motifs: HCG, GCH and GCG. You can change this directly in 02_Megalodon.sh according to the information in varSettings.sh.
• To keep the efficient arraying of SLURM jobs without needing to manually change the number of SLURM arrays in 3 scripts (01b_Demux_Guppy.sh, 02_Megalodon.sh and 03_BigWig_metPlot.sh), these 3 scripts will run on 24 (max. number of barcodes for non-bacterial samples) jobs by default, the jobs with no correct barcodes will terminate immediately with no output.