Last updated: 2017-12-04

Code version: a63bb5f

Bash script

Split genome into 200bp windows and run the coverage command:

/project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed

/project2/gilad/briana/Net-seq/ref_genes/windows_200

make window_200_cov.sh

#!/bin/bash

#SBATCH --job-name=window_200_cov
#SBATCH --time=8:00:00
#SBATCH --partition=broadwl
#SBATCH --mem=50G
#SBATCH --tasks-per-node=4
#SBATCH --mail-type=END



bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18486_dep_chr_sort.bed  > /project2/gilad/briana/Net-seq/ref_genes/windows_200/window_200_cov_18486.txt


bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18508_dep_chr_sort.bed  > /project2/gilad/briana/Net-seq/ref_genes/windows_200/window_200_cov_18508_dep.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18508_nondep_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/windows_200/window_200_cov_18508_nondep.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_19238_dep_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/windows_200/window_200_cov_19238.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed  -b /project2/gilad/briana/Net-seq/data/bed_sort/mayer_SRR1575922_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/windows_200/window_200_cov_mayer.txt 


#bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed   -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed/merged_Net1.chr.bed   > /project2/gilad/briana/Net-seq/ref_genes/windows_200/window_200_cov_merged.txt
#step memory exceeded!

Import data

window_200_18486=read.csv("../data/windows_200/window_200_cov_18486.txt", header=FALSE, sep="\t")
window_200_18508_dep=read.csv("../data/windows_200/window_200_cov_18508_dep.txt", header=FALSE, sep="\t")
window_200_18508_nondep=read.csv("../data/windows_200/window_200_cov_18508_nondep.txt", header=FALSE, sep="\t")
window_200_19238=read.csv("../data/windows_200/window_200_cov_19238.txt", header=FALSE, sep="\t")
window_200_mayer= read.csv("../data/windows_200/window_200_cov_mayer.txt", header=FALSE, sep="\t")

Add col labels to each file:

colnames(window_200_18486) = c("chr", "start", "end", "count")
colnames(window_200_18508_dep) = c("chr", "start", "end", "count")
colnames(window_200_18508_nondep) = c("chr", "start", "end", "count")
colnames(window_200_19238) = c("chr", "start", "end", "count")
colnames(window_200_mayer) = c("chr", "start", "end", "count")

Plot data

Data I want to look at:

  • summary per library
summary(window_200_18486$count)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
      0       0       0       1       0 4076520 
summary(window_200_18508_dep$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
       0        0        0       21        0 27069584 
summary(window_200_18508_nondep$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
       0        0        0       23        0 30571781 
summary(window_200_19238$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
       0        0        0        4        0 13033253 
summary(window_200_mayer$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
     0.0      0.0      0.0    113.3      0.0 701170.0

Use a plot to see the distribution:

  • summaries not including zero

Make dataframe excluding the zeros:

window_200_18486_non0= window_200_18486[window_200_18486$count!=0,]
window_200_18508_dep_non0= window_200_18508_dep[window_200_18508_dep$count!=0,]
window_200_18508_nondep_non0= window_200_18508_nondep[window_200_18508_nondep$count!=0,]
window_200_19238_non0= window_200_19238[window_200_19238$count!=0,]
window_200_mayer_non0= window_200_mayer[window_200_mayer$count!=0,]

summarise

summary(window_200_18486_non0$count)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
      1       1       1      45       3 4076520 
summary(window_200_18508_dep_non0$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
       1        1        1      624        3 27069584 
summary(window_200_18508_nondep_non0$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
       1        1        1      633        3 30571781 
summary(window_200_19238_non0$count)
    Min.  1st Qu.   Median     Mean  3rd Qu.     Max. 
       1        1        1      149        3 13033253 
summary(window_200_mayer_non0$count)
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
      1       1       1    2200       2  701170 
plot(sort(log(window_200_19238_non0$count), decreasing=TRUE))

  • number of entries that are non zero
x= nrow(window_200_18486)
barplot(c(nrow(window_200_18486_non0)/x,nrow(window_200_18508_dep_non0)/x,nrow(window_200_18508_nondep_non0)/x, nrow(window_200_19238_non0)/x), main="Proportion of detected bins", names=c("18486", "18508 \n dep", "18508 \n nondep", "19238"))

nrow(window_200_mayer_non0)/x
[1] 0.05149546

Integrate sets:

window_non_0_all= nrow(window_200_18486[window_200_18486$count!=0 | window_200_18508_dep$count!=0 | window_200_18486$count!=0,] )
prop_window_non0= window_non_0_all/x
prop_window_non0
[1] 0.05416194
window_non_0_2= nrow(window_200_18486[window_200_18486$count!=0 | window_200_18508_dep$count!=0,] )
prop_window_non0_2= window_non_0_2/x
prop_window_non0_2
[1] 0.05416194
window_non_0_2b= nrow(window_200_18486[window_200_18486$count!=0 |window_200_18486$count!=0,] )
prop_window_non0_2b= window_non_0_2b/x
prop_window_non0_2b
[1] 0.03282309

Explore bin overlap

I will combine the depleted samples counts in one data frame:

window_200_3lib= cbind(window_200_18486, window_200_18508_dep$count, window_200_19238$count)
colnames(window_200_3lib)= c("chr", "start", "end", "18486", "18508", "19238")

Make a vector with how many of the libraries have coverage for each bin

sum_vec= apply(window_200_3lib[,4:6],1,function(x)sum(x>=1))
window_200_3lib= cbind(window_200_3lib, sum_vec)

Explore the results:

bin_0= sum(sum_vec==0)
bin_1= sum(sum_vec==1)
bin_2= sum(sum_vec==2)
bin_3= sum(sum_vec==3)

prop0=bin_0/x
prop1=bin_1/x
prop2=bin_2/x
prop3=bin_3/x

barplot_prop=(barplot(c(prop1, prop2, prop3),names = c("1 bin", "2 bins", "3 bins"), main="Proportion of bins with coverage in 3 libaries", ylab="proportion", xlab="library" ))

prop_vec=c(prop0,prop1, prop2, prop3)
names_vec= c("0 bins","1 bin", "2 bins", "3 bins")

prop_table=rbind(names_vec,prop_vec)
prop_table
          [,1]                [,2]                 [,3]                
names_vec "0 bins"            "1 bin"              "2 bins"            
prop_vec  "0.934702546224945" "0.0453430987382321" "0.0115298190107529"
          [,4]                 
names_vec "3 bins"             
prop_vec  "0.00842453602607038"

Perform on mayer data or multiple lanes

#!/bin/bash

#SBATCH --job-name=window_200_cov_mayer
#SBATCH --time=8:00:00
#SBATCH --partition=broadwl
#SBATCH --mem=50G
#SBATCH --tasks-per-node=4
#SBATCH --mail-type=END

module load bedtools

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_200bp_window.bed -b /project2/gilad/briana/mayer.data/mayer_hek/data/bed/mayer_hek-sort.chr.bed > /project2/gilad/briana/mayer.data/mayer_hek/data/window_200_cov_mayer_hek.txt

Not enough memory

Bin the gene file:

bedtools makewindows -b ref_seq_gene_hg19 -w 200 >  ref_seq_gene_hg16_bins.bed

Make a bash script to get coverage in these bins. This is in /project2/gilad/briana/Net-seq/ref_genes/gene_windows_200/gene_window_200.sh

#!/bin/bash

#SBATCH --job-name=gene_window_200_cov
#SBATCH --time=8:00:00
#SBATCH --partition=broadwl
#SBATCH --mem=30G
#SBATCH --tasks-per-node=4
#SBATCH --mail-type=END



bedtools coverage -counts -a /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg16_bins.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18486_dep_chr_sort.bed  > /project2/gilad/briana/Net-seq/ref_genes/gene_windows_200/gene_window_cov_18486.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg16_bins.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18508_dep_chr_sort.bed  > /project2/gilad/briana/Net-seq/ref_genes/gene_windows_200/gene_window_cov_18508_dep.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg16_bins.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_19238_dep_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/gene_windows_200/gene_window_cov_19238.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg16_bins.bed  -b /project2/gilad/briana/Net-seq/data/bed_sort/mayer_SRR1575922_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/gene_windows_200/gene_window_cov_mayer.txt

This file has 22071951 bins. This is not a good representation because some locations in the genome are in multiple genes. I need to think about this more. Maybe I can only keep unique windows.

``ref_seq_gene_hg16_bins.bed | uniq -u | wc -l```

This leaves 5023160 lines. Still not right because the genes start at different positions but overlap then the lines wouldnt be unique.

Use bedtools intersect -a genome -b genes -wa:

  • a is the genome windows

  • b is the gene file

This should keep the windows of a that intersect b. This means I will only have windows that contain a gene.


bedtools intersect -a /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_sort_2.bed -b /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg19_small.bed -wa > /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_with_gene.bed

ERROR:

ERROR: Received illegal bin number 262143 from getBin call.
ERROR: Unable to add record to tree

Potential problem: hg19 file starts with chrom 10 and the gene file starts with chr1. Maybe try sorting the gene file the way I sorted the hg19 file.

Get the chr# list from the genes file using:

cut -f 1 | uniq /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg19_small.bed > names.txt

#!/bin/bash


#SBATCH --job-name=sortgene
#SBATCH --time=8:00:00
#SBATCH --partition=broadwl
#SBATCH --mem=20G
#SBATCH --tasks-per-node=4
#SBATCH --mail-type=END

module load bedtools

cd /project2/gilad/briana/Net-seq/genome_bed
 
bedtools sort -faidx names_uniq.txt  -i  hg19_200bp_window.bed -o hg19_windows_sort.bed

ERROR: Chromosome “chr1” undefined in names_uniq.txt

Try:

sort -k 1,1 -k2,2n hg19_200bp_window.bed > hg19_windows_sort_2.bed

Cut the gene file to only have the first 3 columns with:

cut -f 1,2,3 ref_seq_gene_hg19 > ref_seq_gene_hg19_3col.bed

Run intersect:

bedtools intersect -a /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_sort_2.bed -b /project2/gilad/briana/Net-seq/ref_genes/ref_seq_gene_hg19_3col.bed -wa > /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_with_gene.bed

This worked but you still have multiples. I will take the unique lines of this with:

cat hg19_windows_with_gene.bed | uniq > hg19_windows_with_gene_uniq.bed

Now I can run the coverage counts on this file to see the how many of the genome bins that had at least one gene have coverage.

Call this script. uniq_gene_window.sh in /project2/gilad/briana/Net-seq/ref_genes/uniq_gene_windows_200

#!/bin/bash

#SBATCH --job-name=uniq_gene_window
#SBATCH --time=8:00:00
#SBATCH --partition=broadwl
#SBATCH --mem=30G
#SBATCH --tasks-per-node=4
#SBATCH --mail-type=END



bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_with_gene_uniq.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18486_dep_chr_sort.bed  > /project2/gilad/briana/Net-seq/ref_genes/uniq_gene_windows_200/uniq_gene_window_cov_18486.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_with_gene_uniq.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_18508_dep_chr_sort.bed  > /project2/gilad/briana/Net-seq/ref_genes/uniq_gene_windows_200/uniq_gene_window_cov_18508_dep.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_with_gene_uniq.bed -b /project2/gilad/briana/Net-seq/Net-seq1/data/bed_sort/net1_19238_dep_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/uniq_gene_windows_200/uniq_gene_window_cov_19238.txt

bedtools coverage -counts -a /project2/gilad/briana/Net-seq/genome_bed/hg19_windows_with_gene_uniq.bed -b /project2/gilad/briana/Net-seq/data/bed_sort/mayer_SRR1575922_chr_sort.bed   > /project2/gilad/briana/Net-seq/ref_genes/uniq_gene_windows_200/uniq_gene_window_cov_mayer.txt

Load uniq gene files:

These are the uniq bind in the genome 200bp window file that include a gene.

uniq_gene_window_18486=read.csv("../data/uniq_genes/uniq_gene_window_cov_18486.txt", header=FALSE, sep="\t")
uniq_gene_window_18508=read.csv("../data/uniq_genes/uniq_gene_window_cov_18508_dep.txt", header=FALSE, sep="\t")
uniq_gene_window_19238=read.csv("../data/uniq_genes/uniq_gene_window_cov_19238.txt", header=FALSE, sep="\t")
uniq_gene_window_mayer=read.csv("../data/uniq_genes/uniq_gene_window_cov_mayer.txt", header=FALSE, sep="\t")
colnames(uniq_gene_window_mayer)= c("chr", "start", "end", "count")
colnames(uniq_gene_window_19238)= c("chr", "start", "end", "count")
colnames(uniq_gene_window_18486)= c("chr", "start", "end", "count")
colnames(uniq_gene_window_18508)= c("chr", "start", "end", "count")

Look at the number of windows with coverage:

uniq_gene_window_18486_no0= uniq_gene_window_18486[uniq_gene_window_18486$count!=0,]
uniq_gene_window_18508_no0= uniq_gene_window_18508[uniq_gene_window_18508$count!=0,]
uniq_gene_window_19238_no0= uniq_gene_window_19238[uniq_gene_window_19238$count!=0,]
uniq_gene_window_mayer_no0= uniq_gene_window_mayer[uniq_gene_window_mayer$count!=0,]

Bar plot for coverage:

gene_windows= nrow(uniq_gene_window_mayer)

barplot(c(nrow(uniq_gene_window_18486_no0)/gene_windows, nrow(uniq_gene_window_18508_no0)/gene_windows, nrow(uniq_gene_window_19238_no0)/gene_windows, nrow(uniq_gene_window_mayer_no0)/gene_windows), main="Coverage for windows with genes", names= c("18486", "18508", "19238", "Mayer"), xlab="Library")

mayer_gene_coverage=nrow(uniq_gene_window_mayer_no0)/gene_windows

Mayer has more coverage for this parameter. Their library has 0.0954862.

Ours:

  • 18486: 0.0495328

  • 18508: 0.0460654

  • 19238: 0.0346009

Session information

sessionInfo()
R version 3.4.2 (2017-09-28)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS Sierra 10.12.6

Matrix products: default
BLAS: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

loaded via a namespace (and not attached):
 [1] compiler_3.4.2  backports_1.1.1 magrittr_1.5    rprojroot_1.2  
 [5] tools_3.4.2     htmltools_0.3.6 yaml_2.1.14     Rcpp_0.12.13   
 [9] stringi_1.1.5   rmarkdown_1.6   knitr_1.17      git2r_0.19.0   
[13] stringr_1.2.0   digest_0.6.12   evaluate_0.10.1

This R Markdown site was created with workflowr