TAPS data processing pipeline

Author: Kiki Cano-Gamez

Email: [email protected]

Overview

This repository contains a collection of codes to process methylation data using the TET-assisted pyridine borane sequencing (TAPS) method.

These codes are based on the analytical approach previously published in Science Advances (https://www.science.org/doi/10.1126/sciadv.abh0534), and were developped with assistance from Masato Inoue ([email protected]) and Dr Chunxiao-Song.

Repository structure

The codes contained within this repository correspond to the main data processing steps followed. They are written in bash and ordered as follows:

./
 |-- 0_merge-fastq.sh				Merges sequencing data from different lanes for the same sample into a single FASTQ file.
 |-- 1_trim-galore.sh				Adapter trimming and read clipping with TrimGalore! (i.e. cutadapt, followed by FastQC quality assessment)
 |-- 2_bwa-mem.sh				Alignment of reads to the reference genome using bwa mem
 |-- 3_sort_and_markdup.sh			Quality filtering and sorting of mapped reads with samtools, followed by marking of duplicated reads by picard
 |-- 4_mbias-plot.sh				Methylation bias assessment with methylDackel (used to define read clipping parameters)
 |-- 5_methyl-dackel.sh				Calling of methylation events using methylDackel
 |-- 6_make-bigwig.sh				Creates bigWig files based on the bedGraph outpurs from methylDackel
 |-- 7_get-mapping-stats.sh			Computes summary statistics for the performance of the BWA MEM alignment step (e.g. insert sizes, mapping rates, and genome coverage)
 `-- 8_filter-and-flip-methylation-calls.sh	Flips methylation calls to match TAPS chemistry and removes CpGs overlapping centromeres, gaps, ENCODE blacklisted regions, repetitive regions (repeatMAsker) and common SNPs from methylKit files

Supplementary analyses

Each of the subdirectories within this repository contains a collection of scripts used to perform separate pieces of analysis on TAPS data. These comprise:

identity_check:

Scripts used to assess the extent of genotype sharing between sequencing files. These scripts were used to identify any potential sample swaps or cross-contaminations during data generation.

tissue_deconvolution:

Scripts used to estimate the proportional contribution of different tissues to the cfDNA pool. These scripts perform deconvolution based on methylation patterns at CpG sites known to be hyper- or hypomethylated in a tissue specific manner.

fragmentomics_analysis:

Scripts used to recover fragment length and fragment end-motif information for all cfDNA molecules sequenced.

nucleosome_mapping:

Scripts used to calculate windowed protection scores (WPS) and thus identify the likely position of nucleosomes at TSS regions