overlapMolQTL.Rmd

---
title: "Overlap Molecular QTLs"
author: "Briana Mittleman"
date: "10/1/2018"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

I will use this script to overlap the molQTLs found in [Call molQTL](callMolQTLS.html) analysis with the APA QTLs I found using the [transcript level annotations](PeakToGeneAssignment.html) . 

I want to ask if APA QTLs effect other molecular QTLs. The first step is to find the top snp-gene pair. The permuted value is giving me 1 snp for each peak. I need to find the top snp/peak in this file for each gene. I will then test these snps for significance at 10% fdr.  

Overlap: Use the permulted molecular QTL pvalues to find the significant QTLs for each molecular phenotype I tested. Find each of these snps in the APA nominal file. Take the most stignficant pair and multiple the pvalue by the number of peaks the snp is associated with for that same gene. As a baseline for this test I will randomly choose the same number of snps from molecular QTL and test these in the APA nominal files. I can run this for the total and nuclear.

I want to do this for each of the molecular QTLs, therefore it would be best to upload the necessary files then create a script that can take any of them and create the QQplot.  

##Upload Data:  

Library
```{r}
library(workflowr)
library(reshape2)
library(tidyverse)
library(VennDiagram)
library(data.table)
```


Permuted Results from APA:

```{r}
nuclearAPA=read.table("../data/perm_QTL_trans/filtered_APApeaks_merged_allchrom_refseqGenes_pheno_Nuclear_transcript_permResBH.txt", stringsAsFactors = F, header = T)
totalAPA=read.table("../data/perm_QTL_trans/filtered_APApeaks_merged_allchrom_refseqGenes_pheno_Total_transcript_permResBH.txt", stringsAsFactors = F, header=T)  
```


Permuted results for other QTLs  

```{r}

perm_names=c("pid" ,"nvar","shape1" ,"shape2", "dummy","sid" ,"dist","npval", "slope" , "ppval" ,"bpval")
su30=read.table("../data/other_qtls/fastqtl_qqnorm_4su30.fixed.perm.out", stringsAsFactors = F,col.names = perm_names)
su60=read.table("../data/other_qtls/fastqtl_qqnorm_4su60.fixed.perm.out", stringsAsFactors = F,col.names = perm_names)
rna=read.table("../data/other_qtls/fastqtl_qqnorm_RNAseq_phase2.fixed.perm.out", stringsAsFactors = F,col.names = perm_names)
rnaG=read.table("../data/other_qtls/fastqtl_qqnorm_RNAseqGeuvadis.fixed.perm.out", stringsAsFactors = F,col.names = perm_names)
rib=read.table("../data/other_qtls/fastqtl_qqnorm_ribo_phase2.fixed.perm.out", stringsAsFactors = F,col.names = perm_names)
prot=read.table("../data/other_qtls/fastqtl_qqnorm_prot.fixed.perm.out", stringsAsFactors = F,col.names = c("Gene.stable.ID" ,"nvar","shape1" ,"shape2", "dummy","sid" ,"dist","npval", "slope" , "ppval" ,"bpval"))
```

##Create overlap plot  

I will write this in multiple functions and put them together. The first function will take in the permuted results and return the significant snps at a given FDR.  


First step is to take in the mol file and change the names:  

```{r,eval=F}
geneNames=read.table("/project2/gilad/briana/genome_anotation_data/ensemble_to_genename.txt", sep="\t", header=T,stringsAsFactors = F)
 
file_newNames=mol_file %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval")

```


```{r, eval=F}
#returns significant snps given a file and a cutoff 
sigsnp=function(file, cutoff){
  file$bh=p.adjust(file$bpval, method="fdr")
  file_sig=file %>% filter(-log10(bh)> cutoff) %>% select(Gene.name, sid)
  return(file_sig)
}

testsigsnp=sigsnp(rna,1 )

```

Next step is to choose a random subset with the same number of snps as were found significant.  

```{r,eval=F}

#takes the file and the list of sig snps, returns a df with the same number of random snps  
randomsnps=function(file, SigSnpList){
  nsnp=nrow(SigSnpList)
  randomSnpDF= file %>% sample_n(nsnp) %>% arrange(sid) %>% select(Gene.name,sid) 
  return(randomSnpDF)
}
testrandomsnps=randomsnps(rna, testsigsnp)
```

The next step is to filter nuclear file by the snp id and gene. To do this I will join on the snpIDs then group by the snp ids. I should then be able to take the lowest Pvalue from each group and count how many are in each group to multiply by the number of tests. I will practice this with a small set then make the general function.   

```{r,eval=F}
#filter and fix pvals
filt_tot= totalAPA %>% semi_join(testrandomsnps, by=c("Gene.name","sid") %>% group_by(sid) %>% add_tally() %>% ungroup() %>% mutate(corrPval=bpval * n)
#take top snp
filt_tot_top= filt_tot %>% group_by(sid) %>% top_n(-1, corrPval)
```

Make this into a function for the total and nuclear:  

```{r,eval=F}

nom_names=c("peakID", "sid", "dist", "pval", "slope")
#import total nominal
apaTotNom=read_table("/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes_pheno_Total_NomRes.txt", col_names=nom_name, col_types = c(col_character(), col_character(), col_double(), col_double(), col_double()))
#import nuclear nominal
apaNucNom=read_table("/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes_pheno_Nuclear_NomRes.txt", col_names=nom_name, col_types = nom_names c(col_character(), col_character(), col_double(), col_double(), col_double()))


#takes a list of snps and filters the top corrected snp for each one, returns df
top_Total=function(snp_list){
  filt_tot=apaTotNom %>% separate(peakID, into=c("chr", "start", "end", "id"), sep=":") %>% separate(id, into=c("Gene.name", "strand", "peaknum"), sep="_") %>% semi_join(snp_list, by=c("sid", "Gene.name") %>% group_by(sid, Gene.name) %>% add_tally() %>% ungroup() %>% mutate(corrPval=bpval* n)
  filt_tot_top= filt_tot %>% group_by(sid, Gene.name) %>% top_n(-1, corrPval)
  return(filt_tot_top)
}

#same for nuclear:  
top_Nuclear=function(snp_list){
  filt_nuc=apaNucNom  %>% separate(peakID, into=c("chr", "start", "end", "id"), sep=":") %>% separate(id, into=c("Gene.name", "strand", "peaknum"), sep="_") %>% semi_join(snp_list, by=c("sid", "Gene.name") %>% group_by(sid, Gene.name) %>% add_tally() %>% ungroup() %>% mutate(corrPval=bpval* n)
  filt_nuc_top= filt_nuc %>% group_by(sid, Gene.name) %>% top_n(-1, corrPval)
  return(filt_nuc_top)
}

```

In the full script I will run this on the real QTLs and the random snps. 

The next function will make the plots. I will make one that takes the results of the top_total or top_Nuclear snps.  

```{r,eval=F}


#function returns a QQplot when given the results of the top_X functions. One will be the test set (real QTLs) and 1 will be the baseline snps.

makeQQ=function(test, baseline, Mol, Fraction){
  plot=qqplot(-log10(runif(nrow(baseline))), -log10(baseline$corrPval), ylab="Observed", xlab="Expected", main=paste("Overlap QTLs:", Mol, "with APA", Fraction, sep=" "))
  points(sort(-log10(runif(nrow(test)))), sort(-log10(test$corrPval)), col= alpha("Red"))
  abline(0,1)
  return(plot)
}
```
Put these together in a function: I want to give the function the molQTL file and it will make the total and nuclear plots. This means I need to give it the file to write the png files to.  


createOverlapSigMol2APA.R

```{r,eval=F}
#!/bin/rscripts

#this script creates the files for the molQTLs overlap with total and nuclear APA qtl

library(dplyr)
library(tidyr)
library(ggplot2)
library(readr)
library(optparse)

#this script will take the total and nuclear nominal file for a given, then output files to put the total/nuclear/base/test files into, and the mol QTL permuted results 
option_list = list(
    make_option(c("-T", "--file_Total"), action="store", default=NA, type='character',
              help="input nom file total"),
    make_option(c("-N", "--file_Nuclear"), action="store", default=NA, type='character',
              help="input nom file nuclear"),
    make_option(c("-A", "--output_test_total"), action="store", default=NA, type='character', help="output for test set total"),
    make_option(c("-B", "--output_test_nuclear"), action="store", default=NA, type='character', help="output for test set nulear"),
        make_option(c("-C", "--output_base_total"), action="store", default=NA, type='character', help="output for base total"),
            make_option(c("-D", "--output_base_nuclear"), action="store", default=NA, type='character', help="output for baseset nuclear"),
            make_option(c("-M", "--molPhenoQTLperm"), action="store", default=NA, type='character', help="permuter results for molecular pheno")
)


opt_parser <- OptionParser(option_list=option_list)
opt <- parse_args(opt_parser)
print(opt)

nom_name=c("peakID", "sid", "dist", "pval", "slope")


geneNames=read.table("/project2/gilad/briana/genome_anotation_data/ensemble_to_genename.txt", sep="\t", header=T, stringsAsFactors = F)


#function to run per mol QTLs

overlapQTLplot=function(mol_file, cut, optA=opt, nom_nameA=nom_name){

if (mol_file == "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_prot.fixed.perm.out") {
  in_file=read.table(mol_file, col.names = c("Gene.stable.ID", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval"),stringsAsFactors=F)
  file_newNames=in_file %>%  inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval")
} else {
in_file=read.table(mol_file, col.names = c("pid", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval"),stringsAsFactors=F)
file_newNames=in_file %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval")
}
  
    #helper functions  
#returns significant snps given a file and a cutoff 
  sigsnp=function(file, cutoff){
    file$bh=p.adjust(file$bpval, method="fdr")
    file_sig=file %>% filter(-log10(bh)> cutoff) %>% select(Gene.name, sid)
    return(file_sig)
  }
  randomsnps=function(file, SigSnpList){
    nsnp=nrow(SigSnpList)
    randomSnpDF= file %>% sample_n(nsnp) %>% arrange(sid) %>% select(Gene.name,sid) 
    return(randomSnpDF)
  }

#takes a list of snps and filters the top corrected snp for each one, returns df
    top_Total=function(snp_list,optB=optA,nom_name1=nom_nameA){
      apaTotNom=read.table(optB$file_Total, col.names=nom_name1,stringsAsFactors=F)
      filt_tot=apaTotNom %>% separate(peakID, into=c("chr", "start", "end", "id"), sep=":") %>% separate(id, into=c("Gene.name", "strand", "peaknum"), sep="_") %>% semi_join(snp_list, by=c("sid", "Gene.name")) %>% group_by(sid, Gene.name) %>% add_tally() %>% ungroup() %>%  mutate(corrPvalx=pval* n) %>% mutate(corrPval=ifelse(as.numeric(corrPvalx)>1, "1", corrPvalx))
      filt_tot_top= filt_tot %>% group_by(sid, Gene.name) %>% top_n(-1, corrPvalx)
      return(as.data.frame(filt_tot_top))
    }
    #same for nuclear:  
    top_Nuclear=function(snp_list,optC=optA, nom_name2=nom_nameA){
      apaNucNom=read.table(optC$file_Nuclear, col.names=nom_name2,stringsAsFactors=F)
      filt_nuc=apaNucNom  %>% separate(peakID, into=c("chr", "start", "end", "id"), sep=":") %>% separate(id, into=c("Gene.name", "strand", "peaknum"), sep="_") %>% semi_join(snp_list, by=c("sid", "Gene.name")) %>% group_by(sid, Gene.name) %>% add_tally() %>% ungroup() %>% mutate(corrPvalx=pval* n) %>% mutate(corrPval=ifelse(as.numeric(corrPvalx)>1, "1", corrPvalx))
      filt_nuc_top= filt_nuc %>% group_by(sid, Gene.name) %>% top_n(-1, corrPvalx)
      return(as.data.frame(filt_nuc_top))
    }
    TL=sigsnp(file_newNames, cut)
    BL=randomsnps(file_newNames, TL)
    #top snps test and base total
    topT_T=top_Total(TL)
    topT_B=top_Total(BL)
    
    #top snps test and base total
    topN_T=top_Nuclear(TL)
    topN_B=top_Nuclear(BL)
    return(list(TT=topT_T,TB=topT_B, NT=topN_T,NB=topN_B))

}
 
outputFiles=overlapQTLplot(opt$molPhenoQTLperm, 1)

#write tables 
write.table(outputFiles$TT,opt$output_test_total,quote=F,row.names = F, col.names=T )
write.table(outputFiles$TB,opt$output_base_total,quote=F,row.names = F, col.names=T )
write.table(outputFiles$NT,opt$output_test_nuclear,quote=F,row.names = F, col.names=T )
write.table(outputFiles$NB,opt$output_base_nuclear,quote=F,row.names = F, col.names=T )
```
maybe try as.data.frame

Directories for output:  

Run this on chr 13, protein:


test_createOverlapSigMol2APA.sh
```{bash,eval=F}
#!/bin/bash


#SBATCH --job-name=test_createOverlapSigMol2APA
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=test_createOverlapSigMol2APA.out
#SBATCH --error=test_createOverlapSigMol2APA.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END


module load R


Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr13.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr13.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigProteinQTL_overlapAPA_Total.test.chr13.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigProteinQTL_overlapAPA_Nuclear.test.chr13.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigProteinQTL_overlapAPA_Total.base.chr13.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigProteinQTL_overlapAPA_Nuclear.base.chr13.txt" --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_prot.fixed.perm.out"



```

Run this on all individuals and all phenos:  

run_createOverlapSigMol2APA_prot.sh

```{bash, eval=F}
#!/bin/bash


#SBATCH --job-name=run_createOverlapSigMol2APA_prot
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=run_createOverlapSigMol2APA_prot.out
#SBATCH --error=run_createOverlapSigMol2APA_prot.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END


module load R

#protein 
for i in $(seq 1 22)
do
Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigProteinQTL_overlapAPA_Total.test.chr${i}.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigProteinQTL_overlapAPA_Nuclear.test.chr${i}.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigProteinQTL_overlapAPA_Total.base.chr${i}.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigProteinQTL_overlapAPA_Nuclear.base.chr${i}.txt"  --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_prot.fixed.perm.out"
done 

```

run_createOverlapSigMol2APA_4su30.sh
```{bash, eval=F}
#!/bin/bash


#SBATCH --job-name=run_createOverlapSigMol2APA_4su30
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=run_createOverlapSigMol2APA_4su30.out
#SBATCH --error=run_createOverlapSigMol2APA_4su30.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END


module load R
#4su30
for i in $(seq 1 22)
do
Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su30QTL_overlapAPA_Total.test.chr${i}.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su30QTL_overlapAPA_Nuclear.test.chr${i}.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su30QTL_overlapAPA_Total.base.chr${i}.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su30QTL_overlapAPA_Nuclear.base.chr${i}.txt"  --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_4su30.fixed.perm.out"
done 
```

run_createOverlapSigMol2APA_4su60.sh

```{bash, eval=F}
#!/bin/bash


#SBATCH --job-name=run_createOverlapSigMol2APA_4su60
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=run_createOverlapSigMol2APA_4su60.out
#SBATCH --error=run_createOverlapSigMol2APA_4su60.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END


module load R
#4su60
for i in $(seq 1 22)
do
Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su60QTL_overlapAPA_Total.test.chr${i}.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su60QTL_overlapAPA_Nuclear.test.chr${i}.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su60QTL_overlapAPA_Total.base.chr${i}.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su60QTL_overlapAPA_Nuclear.base.chr${i}.txt"  --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_4su60.fixed.perm.out"
done 

```


run_createOverlapSigMol2APA_RNAsG.sh
```{bash, eval=F}
#!/bin/bash


#SBATCH --job-name=run_createOverlapSigMol2APA_RNAsG
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=run_createOverlapSigMol2APA_RNAsG.out
#SBATCH --error=run_createOverlapSigMol2APA_RNAsG.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END


module load R 
#RNAseqGeuvadis
for i in $(seq 1 22)
do
Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Total.test.chr${i}.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.test.chr${i}.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqGeuvadisQTL_overlapAPA_Total.base.chr${i}.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.base.chr${i}.txt"  --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_RNAseqGeuvadis.fixed.perm.out"
done 


```




run_createOverlapSigMol2APA_RNA.sh
```{bash, eval=F}
#!/bin/bash


#SBATCH --job-name=run_createOverlapSigMol2APA_RNA
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=run_createOverlapSigMol2APA_RNA.out
#SBATCH --error=run_createOverlapSigMol2APA_RNA.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END


module load R

#RNAseq
for i in $(seq 1 22)
do
Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNASeqQTL_overlapAPA_Total.test.chr${i}.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqQTL_overlapAPA_Nuclear.test.chr${i}.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqQTL_overlapAPA_Total.base.chr${i}.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqQTL_overlapAPA_Nuclear.base.chr${i}.txt"  --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_RNAseq_phase2.fixed.perm.out"
done 
```




run_createOverlapSigMol2APA_Ribo.sh
```{bash, eval=F}
#!/bin/bash


#SBATCH --job-name=run_createOverlapSigMol2APA_Ribo
#SBATCH --account=pi-yangili1
#SBATCH --time=24:00:00
#SBATCH --output=run_createOverlapSigMol2APA_Ribo.out
#SBATCH --error=run_createOverlapSigMol2APA_Ribo.err
#SBATCH --partition=bigmem2
#SBATCH --mem=64G
#SBATCH --mail-type=END

module load R

#Ribo
for i in $(seq 1 22)
do
Rscript createOverlapSigMol2APA.R --file_Total "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Total.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --file_Nuclear "/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_trans/filtered_APApeaks_merged_allchrom_refseqGenes.Transcript_sm_quant.Nuclear.pheno_fixed.txt.gz.qqnorm_chr${i}.nominal.out" --output_test_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigriboQTL_overlapAPA_Total.test.chr${i}.txt" --output_test_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigriboQTL_overlapAPA_Nuclear.test.chr${i}.txt" --output_base_total "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigriboQTL_overlapAPA_Total.base.chr${i}.txt" --output_base_nuclear "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigriboQTL_overlapAPA_Nuclear.base.chr${i}.txt"  --molPhenoQTLperm "/project2/gilad/briana/threeprimeseq/data/molecular_QTLs/perm/fastqtl_qqnorm_ribo_phase2.fixed.perm.out"
done 

```


I will need to concatinate all of the test and base files for each phenotype befre running the code to create the QQ plots. 


```{bash,eval=F}
#in base
for i in $(ls *)
do
awk '{if (NR!=1) {print}}' ${i} > ../base_nohead/${i}
done

#in test
for i in $(ls *)
do
awk '{if (NR!=1) {print}}' ${i} > ../test_nohead/${i}
done
```



The results are not large. I will transfer them to my computer and run the analysis here.  


```{r}
makeQQ=function(test, baseline, Mol, Fraction){
  names=c("chr", "start", "end", "Gene.name", "strand" ,"peaknum", "sid", "dist", "pval", "slope", "n" ,"corrPvalX", "corrPval")
  t=read.table(test,stringsAsFactors = F, col.names = names)
  t$corrPval=as.numeric(t$corrPval)
  b=read.table(baseline,stringsAsFactors = F,col.names = names)
  b$corrPval=as.numeric(b$corrPval)
  plot=qqplot(-log10(runif(nrow(b))), -log10(b$corrPval), ylab="Observed", xlab="Expected", main=paste("Overlap QTLs:", Mol, "with APA", Fraction, sep=" "))
      points(sort(-log10(runif(nrow(t)))), sort(-log10(t$corrPval)), col= alpha("Red"))
      abline(0,1)
    return(plot)
}
```

####RNA   
```{r}
rna_total_plot=makeQQ("../data/mol_overlap/test/SigRNAseqQTL_overlapAPA_Total.test.txt", "../data/mol_overlap/base/SigRNAseqQTL_overlapAPA_Total.base.txt", "RNAseq", "Total")
rna_nuc_plot=makeQQ("../data/mol_overlap/test/SigRNAseqQTL_overlapAPA_Nuclear.test.txt", "../data/mol_overlap/base/SigRNAseqQTL_overlapAPA_Nuclear.base.txt", "RNAseq", "Nuclear")

```

###RNA G  


```{r}
rna_total_plot=makeQQ("../data/mol_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Total.test.txt", "../data/mol_overlap/base/SigRNAseqGeuvadisQTL_overlapAPA_Total.base.txt", "RNAseq Geu", "Total")
rna_nuc_plot=makeQQ("../data/mol_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.test.txt", "../data/mol_overlap/base/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.base.txt", "RNAseq Geu", "Nuclear")

```

###4su30   


```{r}
rna_total_plot=makeQQ("../data/mol_overlap/test/Sig4su30QTL_overlapAPA_Total.test.txt", "../data/mol_overlap/base/Sig4su30QTL_overlapAPA_Total.base.txt", "4su 30", "Total")
rna_nuc_plot=makeQQ("../data/mol_overlap/test/Sig4su30QTL_overlapAPA_Nuclear.test.txt", "../data/mol_overlap/base/Sig4su30QTL_overlapAPA_Nuclear.base.txt", "4su 30", "Nuclear")

```


###4su60   


```{r}
rna_total_plot=makeQQ("../data/mol_overlap/test/Sig4su60QTL_overlapAPA_Total.test.txt", "../data/mol_overlap/base/Sig4su60QTL_overlapAPA_Total.base.txt", "4su 60", "Total")
rna_nuc_plot=makeQQ("../data/mol_overlap/test/Sig4su60QTL_overlapAPA_Nuclear.test.txt", "../data/mol_overlap/base/Sig4su60QTL_overlapAPA_Nuclear.base.txt", "4su 60", "Nuclear")

```
###Protein 


```{r}
rna_total_plot=makeQQ("../data/mol_overlap/test/SigProteinQTL_overlapAPA_Total.test.txt", "../data/mol_overlap/base/SigProteinQTL_overlapAPA_Total.base.txt", "Protein", "Total")
rna_nuc_plot=makeQQ("../data/mol_overlap/test/SigProteinQTL_overlapAPA_Nuclear.test.txt", "../data/mol_overlap/base/SigProteinQTL_overlapAPA_Nuclear.base.txt", "Protein", "Nuclear")

```
###Ribo


```{r}
rna_total_plot=makeQQ("../data/mol_overlap/test/SigriboQTL_overlapAPA_Total.test.txt", "../data/mol_overlap/base/SigriboQTL_overlapAPA_Total.base.txt", "Ribo", "Total")
rna_nuc_plot=makeQQ("../data/mol_overlap/test/SigriboQTL_overlapAPA_Nuclear.test.txt", "../data/mol_overlap/base/SigriboQTL_overlapAPA_Nuclear.base.txt", "Ribo", "Nuclear")

```

I want to plot the proportion of QTLs that overlap with APA per pheno type:  

Fix names and look at number of significant:  

```{r}

geneNames=read.table("../data/ensemble_to_genename.txt", sep="\t", header=T,stringsAsFactors = F)

#su30
su30.name=su30 %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID")
su30.name$bh=p.adjust(su30.name$bpval, method="fdr")
su30.name_sig=su30.name %>% filter(-log10(bh)> 1) %>% nrow()
su30.name_sig

su60.name=su60 %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID")
su60.name$bh=p.adjust(su60.name$bpval, method="fdr")
su60.name_sig=su60.name %>% filter(-log10(bh)> 1) %>% nrow()
su60.name_sig

rna.name=rna %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID")
rna.name$bh=p.adjust(rna.name$bpval, method="fdr")
rna.name_sig=rna.name %>% filter(-log10(bh)> 1) %>% nrow()
rna.name_sig

rnaG.name=rnaG %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID")
rnaG.name$bh=p.adjust(rnaG.name$bpval, method="fdr")
rnaG.name_sig=rnaG.name %>% filter(-log10(bh)> 1) %>% nrow()
rnaG.name_sig


rib.name=rib %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID")
rib.name$bh=p.adjust(rib.name$bpval, method="fdr")
rib.name_sig=rib.name %>% filter(-log10(bh)> 1) %>% nrow()
rib.name_sig


prot.name=prot %>% inner_join(geneNames, by="Gene.stable.ID")
prot.name$bh=p.adjust(prot.name$bpval, method="fdr")
prot.name_sig=prot.name %>% filter(-log10(bh)> 1) %>% nrow()
prot.name_sig



phenos=c("4su30", "4su60", "RNA", "RNAG", "Ribo", "Protein")
sig=c(su30.name_sig, su60.name_sig, rna.name_sig, rnaG.name_sig, rib.name_sig, prot.name_sig)

```
Get the Total overlap numbers 
```{r}
names=c("chr", "start", "end", "Gene.name", "strand" ,"peaknum", "sid", "dist", "pval", "slope", "n" ,"corrPvalX", "corrPval")


su30_overT=fread("../data/mol_overlap/test/Sig4su30QTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()
su60_overT=fread("../data/mol_overlap/test/Sig4su60QTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

rna_overT=fread("../data/mol_overlap/test/SigRNAseqQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

rnaG_overT=fread("../data/mol_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

ribo_overT=fread("../data/mol_overlap/test/SigriboQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

prot_overT=fread("../data/mol_overlap/test/SigProteinQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()


overlap_total=c(su30_overT,su60_overT, rna_overT,rnaG_overT,ribo_overT,prot_overT)
```
Get nuclear overlap  
```{r}


su30_overN=fread("../data/mol_overlap/test/Sig4su30QTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()
su60_overN=fread("../data/mol_overlap/test/Sig4su60QTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

rna_overN=fread("../data/mol_overlap/test/SigRNAseqQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

rnaG_overN=fread("../data/mol_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

ribo_overN=fread("../data/mol_overlap/test/SigriboQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()

prot_overN=fread("../data/mol_overlap/test/SigProteinQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) %>% nrow()


overlap_nuclear=c(su30_overN,su60_overN, rna_overN,rnaG_overN,ribo_overN,prot_overN)
```

Make this a dataframe:  

```{r}
overlapDF=as.data.frame(cbind(phenos,sig, overlap_total, overlap_nuclear)) 

overlapDF$sig=as.numeric(as.character(overlapDF$sig))
overlapDF$overlap_total=as.numeric(as.character(overlapDF$overlap_total))
overlapDF$overlap_nuclear=as.numeric(as.character(overlapDF$overlap_nuclear))
overlapDF=overlapDF%>% mutate(Total=overlap_total/sig) %>% mutate(Nuclear=(overlap_nuclear/sig)) %>% dplyr::select(phenos, Total, Nuclear)


overlapDF_melt=melt(overlapDF,id.vars="phenos",variable.name = "Fraction", 
  value.name = "Percent_QTL_Overlap")
```


```{r}
molQTLshare=ggplot(overlapDF_melt, aes(x=phenos, y=Percent_QTL_Overlap, by=Fraction, fill=Fraction)) +geom_bar(stat="identity", position="dodge") + scale_fill_manual(values=c("#5D478B", "#87CEFF")) + labs(title="Percent of Molecular QTLs sharing an APAqtl", x="Molecular Phenotype", y="Percent QTLs at FDR 10%")
ggsave("../output/plots/PercOverlapMolQTL.png", molQTLshare)
```

This is not quite right. This is if I tested it in both. I need to look and see if we have a significant snp gene pair. 


Historgram of the Pvalues from the other significant snp in each phenotype:  
First for the total fraction
```{r}



su30_APApval=read.table("../data/mol_overlap/test/Sig4su30QTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) 
su60_APApval=read.table("../data/mol_overlap/test/Sig4su60QTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) 
rna_APApval=read.table("../data/mol_overlap/test/SigRNAseqQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) 
rnaG_APApval=read.table("../data/mol_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) 
ribo_APApval=read.table("../data/mol_overlap/test/SigriboQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) 
prot_APApval=read.table("../data/mol_overlap/test/SigProteinQTL_overlapAPA_Total.test.txt",stringsAsFactors = F, col.names = names) 
```

Plot a histogram with the corrected Pvalues for each of these:  

```{r}
png("../output/plots/AllPheno_histPval_TotAPA.png",width=600, height=400)
par(mfrow=c(2,3))
hist(su30_APApval$corrPval, breaks=50, main="4su30 (Transcription)", xlab="APA total Pval")
hist(su60_APApval$corrPval, breaks=50,main="4su60 (Transcription)", xlab="APA total Pval")
hist(rna_APApval$corrPval, breaks=50, main= "RNA", xlab="APA total Pval")
hist(rnaG_APApval$corrPval, breaks=50, main="RNA Guevadis", xlab="APA total Pval")
hist(ribo_APApval$corrPval, breaks=50, main="Ribosome (Translation)", xlab="APA total Pval")
hist(prot_APApval$corrPval, breaks=50, main="Protein", xlab="APA total Pval")
dev.off()
```
```{r}


su30_APApvalN=read.table("../data/mol_overlap/test/Sig4su30QTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) 
su60_APApvalN=read.table("../data/mol_overlap/test/Sig4su60QTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) 
rna_APApvalN=read.table("../data/mol_overlap/test/SigRNAseqQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) 
rnaG_APApvalN=read.table("../data/mol_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) 
ribo_APApvalN=read.table("../data/mol_overlap/test/SigriboQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) 
prot_APApvalN=read.table("../data/mol_overlap/test/SigProteinQTL_overlapAPA_Nuclear.test.txt",stringsAsFactors = F, col.names = names) 

```


```{r}
png("../output/plots/AllPheno_histPval_NucAPA.png", width=600, height=400)
par(mfrow=c(2,3))
hist(su30_APApvalN$corrPval, breaks=50, main="4su30", xlab="APA Nuclear Pval")
hist(su60_APApvalN$corrPval, breaks=50,main="4su60", xlab="APA Nuclear Pval")
hist(rna_APApvalN$corrPval, breaks=50, main= "RNA", xlab="APA Nuclear Pval")
hist(rnaG_APApvalN$corrPval, breaks=50, main="RNA Guevadis", xlab="APA Nuclear Pval")
hist(ribo_APApvalN$corrPval, breaks=50, main="Ribosome", xlab="APA Nuclear Pval")
hist(prot_APApvalN$corrPval, breaks=50, main="Protein", xlab="APA Nuclear Pval")
dev.off()
```


###Did not use this on terminal 

I will need to concatinate all of the test and base files for each phenotype befre running the code to create the QQ plots.  

Make QQplots  
```{r,eval=F}

makeQQ=function(test, baseline, Mol, Fraction, plot_name){
  t=read.table(test,stringsAsFactors = F, header=T)
  b=read.table(baseline,stringsAsFactors = F,header=T)
  png(plot_name)    
  plot=qqplot(-log10(runif(nrow(b))), -log10(b$corrPval), ylab="Observed", xlab="Expected", main=paste("Overlap QTLs:", Mol, "with APA", Fraction, sep=" "))
      points(sort(-log10(runif(nrow(t)))), sort(-log10(t$corrPval)), col= alpha("Red"))
      abline(0,1)
    dev.off
}

#run function on each pheno  
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su30QTL_overlapAPA_Total.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su30QTL_overlapAPA_Total.base.txt", "4su 30", "Total", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlap4su30_Total.png")
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su30QTL_overlapAPA_Nuclear.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su30QTL_overlapAPA_Nuclear.base.txt", "4su 30", "Nuclear", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlap4su30_Nuclear.png")

makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su60QTL_overlapAPA_Total.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su60QTL_overlapAPA_Total.base.txt", "4su 60", "Total", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlap4su60_Total.png")
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/Sig4su60QTL_overlapAPA_Nuclear.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/Sig4su60QTL_overlapAPA_Nuclear.base.txt", "4su 60", "Nuclear", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlap4su60_Nuclear.png")

makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Total.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqGeuvadisQTL_overlapAPA_Total.base.txt", "RNAseq_Guevadis", "Total", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapRNAGue_Total.png")
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqGeuvadisQTL_overlapAPA_Nuclear.base.txt", "RNAseq_Guevadis", "Nuclear", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapRNAGue_Nuclear.png")


makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqQTL_overlapAPA_Total.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqQTL_overlapAPA_Total.base.txt", "RNAseq", "Total", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapRNA_Total.png")
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigRNAseqQTL_overlapAPA_Nuclear.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigRNAseqQTL_overlapAPA_Nuclear.base.txt", "RNAseq", "Nuclear", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapRNA_Nuclear.png")

makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigriboQTL_overlapAPA_Total.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigriboQTL_overlapAPA_Total.base.txt", "Ribo Seq", "Total", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapRibo_Total.png")
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigriboQTL_overlapAPA_Nuclear.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigriboQTL_overlapAPA_Nuclear.base.txt", "Ribo Seq", "Nuclear", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapRibo_Nuclear.png")


makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigProteinQTL_overlapAPA_Total.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigProteinQTL_overlapAPA_Total.base.txt", "Protein", "Total", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapProtein_Total.png")
makeQQ("/project2/gilad/briana/threeprimeseq/data/molecular_overlap/test/SigProteinQTL_overlapAPA_Nuclear.test.txt", "/project2/gilad/briana/threeprimeseq/data/molecular_overlap/base/SigProteinQTL_overlapAPA_Nuclear.base.txt", "Protein", "Nuclear", "/project2/gilad/briana/threeprimeseq/output/plots/QTL_overlap/APAoverlapProtein_Nuclear.png")

```






###Change gene IDs  
**test for name changes ** 
http://useast.ensembl.org/biomart/martview/79ab6e7b92009b7da70ce6306b5efb93


```{r}
#geneNames=read.table("../data/ensemble_to_genename.txt", sep="\t", header=T,stringsAsFactors = F)
```

```{r}
#genesAPA=nuclearAPA %>% separate(pid, into=c("chr", "start", "end", "id"), sep=":") %>% separate(id, into=c("Gene.name", "strand", "peaknum"), sep="_") %>% dplyr::select(Gene.name) %>% distinct(Gene.name)  %>% distinct()

```

Make a full list of the genes used in all of the mol phenoytpe files and use this as the input for the getBM.  

```{r}


 
#su30_geneID=su30 %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval") %>% distinct()

#su60_geneID=su60 %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval") %>% distinct()


#rna_geneID=rna %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval") %>% distinct()


#rib_geneID=rib %>% separate(pid, into=c("Gene.stable.ID", "ver"), sep ="[.]") %>% inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval") %>% distinct()


#prot_geneID=prot %>% rename("Gene.stable.ID"=pid) %>%  inner_join(geneNames, by="Gene.stable.ID") %>% dplyr::select("Gene.name", "nvar", "shape1", "shape2", "dummy", "sid", "dist", "npval", "slope", "ppval", "bpval") %>% distinct()



```