In musikutiv/tsTools: Toolbox for Functional Genomics Data Processing in R

Alignment of MNaseSeq-Data

Paired-end data is aligned as follows using bowtie2.

• only concordant reads are reported
• maximum insert size is kept to 250 (a bit larger than nucleosomes)
• multiple matches of to the genome are suppressed, orphaned mates are eliminated
• BAM file is then converted to BED in which chr, start, end of the sequenced fragement is reported

Cleaning up of orphaned reads is performed with this script taken from (https://www.biostars.org/p/95929/) ```{python, eval = FALSE,python.reticulate = FALSE}

!/usr/bin/env python

import csv import sys

f = csv.reader(sys.stdin, dialect="excel-tab") of = csv.writer(sys.stdout, dialect="excel-tab") last_read = None for line in f : #take care of the header if(line[0][0] == "@") : of.writerow(line) continue

if(last_read == None) : 
    last_read = line
else :
    if(last_read[0] == line[0]) :
        of.writerow(last_read)
        of.writerow(line)
        last_read = None
    else :
        last_read = line

```{bash, eval=FALSE}
BOWTIE_INDEX= <dir>/Bowtie2Index/genome
BOWTIE_OPTS="-p 10 -X 250 --no-discordant --no-mixed --no-unal"
bowtie2 $BOWTIE_OPTS -x $BOWTIE_INDEX -1 mate_1.fastq.gz -2 mate_2.fastq.gz > aligned.sam
samtools view -hf 0x2 aligned.sam | grep -v "XS:i:" | filter_orphans.py | samtools view -b -o aligned.bam
samtools sort -n -m 1G -@ 8 -o aligned.s.bam aligned.bam
bamToBed -i aligned.s.bam -bedpe > aligned.bed 2>/dev/null
cut -f 1,2,6 aligned.bed > aligned.s.bed

Single-read data is aligned as follows using bowtie2.

• multiple matches of to the genome are suppressed
• BAM file is then converted to BED in which chr, start, end and strand of the read is reported

```{bash, eval = FALSE} BOWTIE_INDEX=

/Bowtie2Index/genome BOWTIE_OPTS="-p 10 --no-unal" bowtie2 $bowtie_opts -x $BOWTIE_INDEX ${SRUN}.fastq.gz > aligned.sam samtools view -h aligned.sam | grep -v "XS:i:" | samtools view -b -o aligned.bam samtools sort -m 1G -@ 8 -o aligned.s.bam aligned.bam bamToBed -i aligned.s.bam > aligned.bed 2>/dev/null cut -f 1,2,3,6 aligned.bed > aligned.s.bed

### Conversion to coverage object

```r
library(tsTools)
library(IRanges)
library(GenomicRanges)

As an example we load the data included in the package (S. cerevisiae MNase-Seq, subset to chromosome IV)

fpath <- system.file("extdata", "SRR2154281_IV.s.bed", package="tsTools")
cov <- bed2dyad(fpath,"PAIRED",1)

Autocorrelation Function

The autocorrelation function on the signal vector can provide a hint as to whether we have something nucleosomal in the data. Given the beads on a string arrangement of nucleosomes we expect a periodic autocorrelation pattern.

acf(as.vector(cov[['IV']]), lag.max = 1000, main="", xlab="lag (bp)")

Cumulative Plots

data(ann)
head(ann)
ann <- ann[ann$chr=="IV",]
centers <- GRanges(ann$chr, IRanges(ifelse(ann$strand=="+", ann$start, ann$end)), strand=ann$strand)
names(centers) <- rownames(ann)
mat <- coverageWindowsCenteredStranded(centers, window.size=2000, cov)
x <- seq(-1000,1000)
plot(x, apply(mat, 2, mean), type="l", 
     xlab="position relative to center position", ylab="dyad density")

row-wise normalization

mat.sq <- norm.square(mat)
mat.sum <- norm.sum(mat)

par(mfrow=c(1,2))
plot(x, apply(mat.sq, 2, mean), type="l", 
     xlab="position relative to center position", ylab="dyad density (normalized)", main="sum of squares normalization")
plot(x, apply(mat.sum, 2, mean), type="l", 
     xlab="position relative to center position", ylab="dyad density (normalized)", main="sum of row normalization")

mat.bin <- bin.matrix.rows(mat.sum,25)
x <- as.integer(colnames(mat.bin))
plot(x, apply(mat.bin, 2, mean), type="l", 
     xlab="position relative to center position", ylab="dyad density (normalized)", main="binned")
library(zoo)
mat.bin.smoothed <- t(apply(mat.bin, 1, function(x) {zoo::rollmean(x, 9)}))
x <- as.integer(colnames(mat.bin.smoothed))
plot(x, apply(mat.bin.smoothed, 2, mean), type="l", 
     xlab="position relative to center position", ylab="dyad density (normalized)", main="binned & smoothed")

Cumulative plot with scatter

cumPlot(mat.sum, base.col = "#FF0000")

Heatmaps

library(grid)
hmat <- meanScale(mat)
plotRasterHeatmap(hmat)

Getting Spacing and Phasing along Reference points

ann[1:10,]
result <- ocampo(cov, ann[1:10,])
result

sessionInfo()

musikutiv/tsTools documentation built on July 25, 2024, 2:42 a.m.