Evaluate Brain Data

Last updated: 2018-07-17

• ✔ R Markdown file: up-to-date

  Great! Since the R Markdown file has been committed to the Git repository, you know the exact version of the code that produced these results.

• ✔ Environment: empty

  Great job! The global environment was empty. Objects defined in the global environment can affect the analysis in your R Markdown file in unknown ways. For reproduciblity it’s best to always run the code in an empty environment.

• ✔ Seed: set.seed(12345)

  The command set.seed(12345) was run prior to running the code in the R Markdown file. Setting a seed ensures that any results that rely on randomness, e.g. subsampling or permutations, are reproducible.

• ✔ Session information: recorded

  Great job! Recording the operating system, R version, and package versions is critical for reproducibility.

• ✔ Repository version: c287f4e

  Great! You are using Git for version control. Tracking code development and connecting the code version to the results is critical for reproducibility. The version displayed above was the version of the Git repository at the time these results were generated.
  
  Note that you need to be careful to ensure that all relevant files for the analysis have been committed to Git prior to generating the results (you can use wflow_publish or wflow_git_commit). workflowr only checks the R Markdown file, but you know if there are other scripts or data files that it depends on. Below is the status of the Git repository when the results were generated:

  
  Ignored files:
      Ignored:    .DS_Store
      Ignored:    .Rhistory
      Ignored:    .Rproj.user/
      Ignored:    output/.DS_Store
  
  Untracked files:
      Untracked:  analysis/test.smash.Rmd
      Untracked:  data/18486.genecov.txt
      Untracked:  data/APApeaksYL.total.inbrain.bed
      Untracked:  data/YL-SP-18486-T_S9_R1_001-genecov.txt
      Untracked:  data/bedgraph_peaks/
      Untracked:  data/bin200.5.T.nuccov.bed
      Untracked:  data/bin200.Anuccov.bed
      Untracked:  data/bin200.nuccov.bed
      Untracked:  data/gene_cov/
      Untracked:  data/leafcutter/
      Untracked:  data/nuc6up/
      Untracked:  data/reads_mapped_three_prime_seq.csv
      Untracked:  data/ssFC200.cov.bed
      Untracked:  output/picard/
      Untracked:  output/plots/
      Untracked:  output/qual.fig2.pdf
  
  Unstaged changes:
      Modified:   analysis/dif.iso.usage.leafcutter.Rmd
      Modified:   analysis/explore.filters.Rmd
      Modified:   analysis/test.max2.Rmd
      Modified:   code/Snakefile
  
  

  Note that any generated files, e.g. HTML, png, CSS, etc., are not included in this status report because it is ok for generated content to have uncommitted changes.

I downloaded the brain 3’ seq data from https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM747470 and I want to use this analysis to see how similar their peaks are to ours eventhough the data is from different cell types.

First I will use the bedtools jaccard function to explore the overlaps. It will give me one stat that is the length(intersection)/length(union) - length(intersection). Here I can have file A brain peaks and file B be our peaks to see the similarity between the sets.

#!/bin/bash

#SBATCH --job-name=jaccard_brain
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=jacard_brain.out
#SBATCH --error=jacard_brain.err
#SBATCH --partition=broadwl
#SBATCH --mem=16G
#SBATCH --mail-type=END



module load Anaconda3
source activate three-prime-env 

bedtools jaccard -a /project2/gilad/briana/threeprimeseq/data/derti_brain/GSM747470_human_brain.sites.clustered.hg19.sort.bed -b /project2/gilad/briana/threeprimeseq/data/peaks/APApeaksYL.total.bed  > /project2/gilad/briana/threeprimeseq/data/derti_brain/total.jaccard.txt 

Results: intersection union-intersection jaccard n_intersections 21371 25414133 0.00084091 21352

The brain set has 89110 peaks and our set has 288350. I will filter ours by score then see if the top 25% have a higher overlap percentage.

library(workflowr)

Loading required package: rmarkdown

This is workflowr version 1.0.1
Run ?workflowr for help getting started

library(dplyr)

Warning: package 'dplyr' was built under R version 3.4.4

Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

library(tidyr)
library(ggplot2)
library(reshape2)

Warning: package 'reshape2' was built under R version 3.4.3

Attaching package: 'reshape2'

The following object is masked from 'package:tidyr':

    smiths

YL_peaks=read.table("../data/bedgraph_peaks/APApeaks.bed", col.names = c("chr", "start", "end", "count", "strand", "score")) %>% mutate(length=end-start)

Warning: package 'bindrcpp' was built under R version 3.4.4

I want the counts for the top 25% of the peaks.

quantile(YL_peaks$count)

          0%          25%          50%          75%         100% 
1.000000e+00 1.343902e+01 2.353933e+01 6.091061e+01 1.604636e+06 

I will subset the peaks by having a count > 61.

awk '$4 >= 60 {print}' APApeaksYL.total.bed > APApeaksYL.top25.total.bed

I can rerun the jaccard with this and see if it changes, this new file has 72877 peaks.

Results:
intersection union-intersection jaccard n_intersections 13221 6452066 0.00204911 13207

The proportion of overlap increased. Next I can try to make plots where I seperate my peaks by if they have a corresponding one in the brain file then plot the scores. To do this I will first use bedtool intersect to get just my peaks that contain a peak in the brain file. I can then use dplyr to merge them.

Here A is my file and B is the brain file.

#!/bin/bash

#SBATCH --job-name=int_brain
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=int.brain.out
#SBATCH --error=int.brain.err
#SBATCH --partition=broadwl
#SBATCH --mem=16G
#SBATCH --mail-type=END


module load Anaconda3
source activate three-prime-env 

bedtools intersect -wa -a /project2/gilad/briana/threeprimeseq/data/peaks/APApeaksYL.total.bed  -b /project2/gilad/briana/threeprimeseq/data/derti_brain/GSM747470_human_brain.sites.clustered.hg19.sort.bed > /project2/gilad/briana/threeprimeseq/data/derti_brain/APApeaksYL.total.inbrain.bed 

The resulting file has 21378 peaks.

YL_peaks_overlap=read.table("../data/APApeaksYL.total.inbrain.bed", col.names = c("chr", "start", "end", "count", "strand", "score")) %>% mutate(length=end-start) %>% mutate(in_brain="Y")

Now I need to join them.

YL_peaks_join=YL_peaks %>% full_join(YL_peaks_overlap, by = c("chr", "start", "end", "count", "strand", "score", "length"))

YL_peaks_join$in_brain[is.na(YL_peaks_join$in_brain)]="N"

YL_peaks_join_sel=YL_peaks_join %>% select(count, in_brain)

Plot these.

ggplot(YL_peaks_join_sel, aes(y=log10(count), x=in_brain, fill=in_brain)) + geom_boxplot() + labs(x="Peak called in brain dataset", y="log10 Score", title="Peak score distribution by inclusion in brain dataset")

ggplot(YL_peaks_join_sel, aes(x=log10(count), fill=in_brain), bins=50) + geom_density(position="identity", alpha=.5) + labs(x="log10 of Score", title="Distribution of log10 Scores in peaks included in brain dataset")

It would be better if the background was just a random subset of the same number. There are 21378 included peaks so I should select a random 21378 to make a background distribution.

samp_YLpeaks= sample_n(YL_peaks, 21378)

ggplot() + geom_histogram(data=samp_YLpeaks, aes(log10(count)), bins=100) + geom_histogram(data=YL_peaks_overlap, aes(log10(count)),fill="Red", bins=100) + labs(x="Log10 of Score", title="Scores in Overlapping set compared to scores in random set")

The next step is to download the Brain fastq data and call peaks using Yangs script. I used sra-tools to download SRR299106. Then I ran my snakemake pipeline on it.

I need to make the genome cov file then use Yangs script to call the peaks.

#!/bin/bash

#SBATCH --job-name=braingencovsplit
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=brain_gencovsplit.out
#SBATCH --error=brain_gencovaplit.err
#SBATCH --partition=broadwl
#SBATCH --mem=40G
#SBATCH --mail-type=END


module load Anaconda3
source activate three-prime-env 


bedtools genomecov -ibam /project2/gilad/briana/derti_brain_raw/data/sort/derti_brain-sort.bam  -d -split > /project2/gilad/briana/derti_brain_raw/data/gencov/derti_brain.gencov.bed

Wrap Yangs script:

#!/bin/bash

#SBATCH --job-name=w_getpeakYLB
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=w_getpeakYLB.out
#SBATCH --error=w_getpeakYLB.err
#SBATCH --partition=broadwl
#SBATCH --mem=12G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


for i in $(seq 1 22); do 
  sbatch callPeaksYL_derti.py $i
done

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     

other attached packages:
[1] bindrcpp_0.2.2  reshape2_1.4.3  ggplot2_2.2.1   tidyr_0.7.2    
[5] dplyr_0.7.5     workflowr_1.0.1 rmarkdown_1.8.5

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.17      compiler_3.4.2    pillar_1.1.0     
 [4] git2r_0.21.0      plyr_1.8.4        bindr_0.1.1      
 [7] R.methodsS3_1.7.1 R.utils_2.6.0     tools_3.4.2      
[10] digest_0.6.15     evaluate_0.10.1   tibble_1.4.2     
[13] gtable_0.2.0      pkgconfig_2.0.1   rlang_0.2.1      
[16] yaml_2.1.19       stringr_1.3.1     knitr_1.18       
[19] rprojroot_1.3-2   grid_3.4.2        tidyselect_0.2.4 
[22] glue_1.2.0        R6_2.2.2          purrr_0.2.5      
[25] magrittr_1.5      whisker_0.3-2     backports_1.1.2  
[28] scales_0.5.0      htmltools_0.3.6   assertthat_0.2.0 
[31] colorspace_1.3-2  labeling_0.3      stringi_1.2.2    
[34] lazyeval_0.2.1    munsell_0.4.3     R.oo_1.22.0      

This reproducible R Markdown analysis was created with workflowr 1.0.1