Empirical Distribution for overlap

Step 1: fix names for eQTL files:

Eqtl files:

• Perm res: /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.perm.out

• Now res: /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.nominal.out

Annotation data= /project2/gilad/briana/genome_anotation_data/ensemble_to_genename.txt

Make file with sig eQTLs:

Run this interactviely in R

library(dplyr)

names=c('pid', 'nvar', 'shape1', 'shape2','dummy', 'sid', 'dist', 'npval' ,'slope', 'ppval', 'bpval' )
permRes=read.table("/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.perm.out",stringsAsFactors = F, col.names = names)

permRes$bh=p.adjust(permRes$bpval, method="fdr")


permRes_sig=permRes %>% filter(-log10(bh)>=1)
nonSig=permRes %>% filter(-log10(bh)<1)


write.table(permRes_sig, file="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_significanteQTLs.txt", col.names=F, row.names=F, quote=F, sep="\t")

write.table(nonSig, file="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs.txt", col.names=F, row.names=F, quote=F, sep="\t")

Make a script to switch the gene names:

changeENSG2GeneName.py

def main(symbolF, genenameF):
  nameFile=open("/project2/gilad/briana/genome_anotation_data/ensemble_to_genename.txt","r")
  name_dic={}
  for i,ln in enumerate(nameFile):
      if i >0:
          id, name, source =ln.split("\t")
          name_dic[id]=name
  inFile=open(symbolF, "r")
  outFile=open(genenameF, "w")
  nnoGene=0
  for ln in inFile:
      geneID=ln.split()[0].split(".")[0]
      #print(geneID)
      if geneID in name_dic.keys():
          geneName=name_dic[geneID]
      else:
          nnoGene+=1
          continue

      otherStuff=ln.split()[1:]
      line=[geneName,  "\t".join(otherStuff)]
      outString="\t".join(line)
      #print(outString)
      outFile.write("%s\n"%(outString))
  print("%s Genes have no associated name in annoation"%(nnoGene))
  outFile.close()
      

if __name__ == "__main__":
    import sys
    ensg=sys.argv[1]
    geneName=sys.argv[2]
    main(ensg, geneName) 

Run for each file:

python changeENSG2GeneName.py  /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.perm.out  /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.perm_GeneNames.out 

python changeENSG2GeneName.py  /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.nominal.out /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/fastqtl_qqnorm_RNAseq_phase2.fixed.nominal_GeneNames.out


python changeENSG2GeneName.py /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_significanteQTLs.txt /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_significanteQTLs_GeneNames.txt

python changeENSG2GeneName.py /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs.txt /project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_eneNames.txt

Step 2: Subset for tested in APA

Subset for genes tested in apa:

I need to subset the non eQTL genes for those actually tested in APA. This is important for the base distribution when I pick a random snp. I will have to do this seperate for total and nuclear similar to how i will have to look at number of peaks tested seperatly in total and nuclear. I will make a dictionary with the genes in the apa analysis, I can then go through the noneqtl file and keep only the genes with a gene in the dictionary.

subsetEQTL_testinAPA.py

totAPA="/project2/gilad/briana/threeprimeseq/data/perm_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno.NoMP_sm_quant.Total.fixed.pheno_5perc_permRes.txt"
NucAPA="/project2/gilad/briana/threeprimeseq/data/perm_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno.NoMP_sm_quant.Nuclear.fixed.pheno_5perc_permRes.txt"


nonEQTLs="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames.txt"
eQTLs="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_significanteQTLs_GeneNames.txt"

eQTLS_intot="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inTot.txt"
eQTLS_innuc="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inNuc.txt"

nonEQTL_intot="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inTot.txt"

nonEQTL_innuc="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inNuc.txt"


def check_names(inApa, ineqtl, outFile):
    gene_dic={}
    
    for ln in open(inApa, "r"):
        gene=ln.split()[0].split(":")[-1].split("_")[0]
        if gene not in gene_dic.keys():
            gene_dic[gene]=""
        else:
            continue
    fileOut= open(outFile, "w")
    nnottested=0
    for ln in open(ineqtl, "r"):
        gene=ln.split()[0]
        if gene in gene_dic.keys():
            fileOut.write(ln)
        else:
            nnottested += 1 
    print("Number of genes not tested in apa = %s"%(nnottested))
    fileOut.close()
    
    
print("total")
check_names(totAPA,nonEQTLs, nonEQTL_intot)
print("nuclear")
check_names(NucAPA,nonEQTLs, nonEQTL_innuc)

print("total")
check_names(totAPA,eQTLs, eQTLS_intot)
print("nuclear")
check_names(NucAPA,eQTLs, eQTLS_innuc)

There are about 4k genes of the 13k not tested.

Step 3: Make a file with # of peaks per gene:

I can make this in R with group by in the permuted data. I want a file that has the gene and how many peaks are in it.

TotPeaks=read.table("../data/perm_QTL_GeneLocAnno_noMP_5percov/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno.NoMP_sm_quant.Total.fixed.pheno_5perc_permResBH.txt", header = T, stringsAsFactors = F) %>% select(pid) %>% separate(pid, into=c("chr", "start", "end", "peak"), sep=":") %>%  separate(peak, into=c("gene", 'strand', 'peaknum'), sep="_") %>% group_by(gene) %>% summarise(NPeaks=n())

Warning: Expected 3 pieces. Additional pieces discarded in 3 rows [886,
887, 888].

write.table(TotPeaks, file="../data/EmpiricalDists/TotalQTL_nPeaks.txt", quote=F, sep="\t", col.names = F, row.names = F)


NucPeaks=read.table("../data/perm_QTL_GeneLocAnno_noMP_5percov/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno.NoMP_sm_quant.Nuclear.fixed.pheno_5perc_permResBH.txt", header = T, stringsAsFactors = F) %>% select(pid) %>% separate(pid, into=c("chr", "start", "end", "peak"), sep=":") %>%  separate(peak, into=c("gene", 'strand', 'peaknum'), sep="_") %>% group_by(gene) %>% summarise(NPeaks=n())

Warning: Expected 3 pieces. Additional pieces discarded in 3 rows [1056,
1057, 1058].

write.table(NucPeaks, file="../data/EmpiricalDists/NuclearQTL_nPeaks.txt", quote=F, sep="\t", col.names = F, row.names = F)

Transfer to /project2/gilad/briana/threeprimeseq/data/EmpiricalDists/

Step 4 Make dist

makeeQTLEmpirical.py

def main(peaks, APA,nonE,eqtl, outF):
    #dictionary of non egenes  
    nonEgenes={}
    for ln in open(nonE):
        gene=ln.split()[0]
        nonEgenes[gene]=""
        
    # make dictionary- key= npeak, value=list of genes with same value (use this when i need to pull random from the values)
    #only add genes in the non eqet
    peaknum={}
    for ln in open(peaks, "r"):
        gene, npeak = ln.split()
        if npeak not in peaknum.keys():
            if gene in nonEgenes.keys():
                peaknum[npeak]=[gene]
        else:
            if gene in nonEgenes.keys():
                 peaknum[npeak].append(gene)
    #make opposite dic- key =gene, value = npeaks (use this when i first get an EGENE to see how many peaks i will need)
    #want all genes- e and no e genes 
    GenePeak={}
    for ln in open(peaks, "r"):
        gene, npeak= ln.split()
        GenePeak[gene]=npeak
    eQTLs= open(eqtl, "r")
    for ln in eQTLs:
        gene= ln.split()[0]
        peaks2use=GenePeak[gene]
        posibleGenes=peaknum[peaks2use]
        if len(posibleGenes) == 0:
             posibleGenes=peaknum[str(int(peaks2use)+1)]
             if len(posibleGenes) == 0:
                 print("No genes with peaknum")
                 continue
        GeneUse=random.choice(posibleGenes)
        print(GeneUse)
        #find the snp that goes with this gene in the non egenes file- this gives me the lowest pval snp
        for ln in open(nonE,"r"):
            genenon=ln.split()[0]
            #print(genenon)
            if genenon == GeneUse:
                snpUse=ln.split()[5]
                print(snpUse)
        #now i have the gene and the snp, i can write out any of the nominal values from the apa file with these
        fout=open(outF,"w")
        for ln in open(APA, "r"):
            geneApa=ln.split()[0].split(":")[-1].split("_")[0]
            snpApa =ln.split()[1]
            if geneApa==GeneUse and snpApa==snpUse: 
                fout.write(ln)
    fout.close()
  
  

if __name__ == "__main__":
    import sys
    import random
    fraction=sys.argv[1]
    OutFile=sys.argv[2]
    if fraction=="Total":
        npeaks="/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/TotalQTL_nPeaks.txt"
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Total.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        nonEfile="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inTot.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inTot.txt"
    else:
        npeaks="/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/NuclearQTL_nPeaks.txt"
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Nuclear.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        nonEfile="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inNuc.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inNuc.txt"
    main(npeaks, nomFile, nonEfile, eQTL, OutFile) 

 python test_makeeQTLEmpirical.py Total "../data/EmpiricalDists/testTotal.txt"

run_makeeQTLEmpirical.sh

#!/bin/bash

#SBATCH --job-name=run_makeeQTLEmpirical
#SBATCH --account=pi-yangili1
#SBATCH --time=36:00:00
#SBATCH --output=run_makeeQTLEmpirical.out
#SBATCH --error=run_makeeQTLEmpirical.err
#SBATCH --partition=broadwl
#SBATCH --mem=36G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


python makeeQTLEmpirical.py "Total" "/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/eQTL_total_EmpiricalDist.txt"

python makeeQTLEmpirical.py "Nuclear" "/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/eQTL_nuclear_EmpiricalDist.txt"

Overlap the eQTLs and apavalues

I want me make a script that takes in the total and nuclear eqtl overlap and nominal pvalues for the true dists.

getApaPval4eQTLs.py

def main(apa, eqtl, outFile):
    fout=open(outFile,"w")
    for ln in open(eQTL,"r"):
        egene=ln.split()[0]
        eSNP=ln.split()[5]
        for ln in open(apa, "r"):
            geneApa=ln.split()[0].split(":")[-1].split("_")[0]
            snpApa =ln.split()[1]   
            if geneApa==egene and snpApa==eSNP:
                fout.write(ln)
    fout.close()
                

if __name__ == "__main__":
    import sys
    import random
    fraction=sys.argv[1]
    OutFile=sys.argv[2]
    if fraction=="Total":
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Total.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inTot.txt"
    else:
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Nuclear.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inNuc.txt"
    main(nomFile, eQTL, OutFile) 
    
    

run_getApaPval4eQTLs.sh

#!/bin/bash

#SBATCH --job-name=run_getApaPval4eQTLs
#SBATCH --account=pi-yangili1
#SBATCH --time=36:00:00
#SBATCH --output=run_getApaPval4eQTLs.out
#SBATCH --error=run_getApaPval4eQTLs.err
#SBATCH --partition=broadwl
#SBATCH --mem=36G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


python getApaPval4eQTLs.py "Total" "/project2/gilad/briana/threeprimeseq/data/eQTL_inAPA/eQTLinTotalApa.txt"

python getApaPval4eQTLs.py "Nuclear" "/project2/gilad/briana/threeprimeseq/data/eQTL_inAPA/eQTLinNuclearApa.txt"

FASTER SCRIPTS:

getApaPval4eQTLs_secondtry.py

def main(apa, eqtl, outFile):
    fout=open(outFile,"w")
    use_dic={}
    for ln in open(eQTL,"r"):
        egene=ln.split()[0]
        eSNP=ln.split()[5]
        use_dic[egene]=eSNP
    for ln in open(apa, "r"):
        geneApa=ln.split()[0].split(":")[-1].split("_")[0]
        snpApa =ln.split()[1]   
        if geneApa in use_dic.keys():
            if snpApa == use_dic[geneApa]:
                fout.write(ln)
    fout.close()
                

if __name__ == "__main__":
    import sys
    import random
    fraction=sys.argv[1]
    OutFile=sys.argv[2]
    if fraction=="Total":
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Total.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inTot.txt"
    else:
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Nuclear.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inNuc.txt"
    main(nomFile, eQTL, OutFile) 
    
    

run_getApaPval4eQTLs_second.sh

#!/bin/bash

#SBATCH --job-name=run_getApaPval4eQTLs2
#SBATCH --account=pi-yangili1
#SBATCH --time=36:00:00
#SBATCH --output=run_getApaPval4eQTLs2.out
#SBATCH --error=run_getApaPval4eQTLs2.err
#SBATCH --partition=broadwl
#SBATCH --mem=36G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


python getApaPval4eQTLs_secondtry.py "Total" "/project2/gilad/briana/threeprimeseq/data/eQTL_inAPA/eQTLinTotalApa_try2.txt"

python getApaPval4eQTLs_secondtry.py "Nuclear" "/project2/gilad/briana/threeprimeseq/data/eQTL_inAPA/eQTLinNuclearApa_try2.txt"

makeeQTLEmpirical_secondtry.py

def main(peaks, APA,nonE,eqtl, outF):
    #dictionary of non egenes  
    nonEgenes={}
    for ln in open(nonE):
        gene=ln.split()[0]
        nonEgenes[gene]=""
        
    # make dictionary- key= npeak, value=list of genes with same value (use this when i need to pull random from the values)
    #only add genes in the non eqet
    peaknum={}
    for ln in open(peaks, "r"):
        gene, npeak = ln.split()
        if npeak not in peaknum.keys():
            if gene in nonEgenes.keys():
                peaknum[npeak]=[gene]
        else:
            if gene in nonEgenes.keys():
                 peaknum[npeak].append(gene)
    #make opposite dic- key =gene, value = npeaks (use this when i first get an EGENE to see how many peaks i will need)
    #want all genes- e and no e genes 
    GenePeak={}
    for ln in open(peaks, "r"):
        gene, npeak= ln.split()
        GenePeak[gene]=npeak
    eQTLs= open(eqtl, "r")
    gene_useDic={}
    for ln in eQTLs:
        gene= ln.split()[0]
        peaks2use=GenePeak[gene]
        posibleGenes=peaknum[peaks2use]
        if len(posibleGenes) == 0:
             print("No genes with peaknum")
             posibleGenes=peaknum[str(int(peaks2use)+1)]
             if len(posibleGenes) == 0:
                 print("No genes with peaknum")
                 continue
        GeneUse=random.choice(posibleGenes)
        gene_useDic[GeneUse]=""
        #print(GeneUse)
        #find the snp that goes with this gene in the non egenes file- this gives me the lowest pval snp
    for ln in open(nonE,"r"):
        genenon=ln.split()[0]
        #print(genenon)
        if genenon in gene_useDic.keys():
            snpUse=ln.split()[5]
            gene_useDic[genenon]=snpUse
        #now i have the gene and the snp, i can write out any of the nominal values from the apa file with these
    fout=open(outF,"w")
    for ln in open(APA, "r"):
        geneApa=ln.split()[0].split(":")[-1].split("_")[0]
        snpApa =ln.split()[1]
        if geneApa in gene_useDic.keys():
            if snpApa == gene_useDic[geneApa]:
                fout.write(ln)
    fout.close()
  
  

if __name__ == "__main__":
    import sys
    import random
    fraction=sys.argv[1]
    OutFile=sys.argv[2]
    if fraction=="Total":
        npeaks="/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/TotalQTL_nPeaks.txt"
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Total.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        nonEfile="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inTot.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inTot.txt"
    else:
        npeaks="/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/NuclearQTL_nPeaks.txt"
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Nuclear.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        nonEfile="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inNuc.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inNuc.txt"
    main(npeaks, nomFile, nonEfile, eQTL, OutFile) 

run_makeeQTLEmpirical_secondtry.sh

#!/bin/bash

#SBATCH --job-name=run_makeeQTLEmpirical2
#SBATCH --account=pi-yangili1
#SBATCH --time=36:00:00
#SBATCH --output=run_makeeQTLEmpirical2.out
#SBATCH --error=run_makeeQTLEmpirical2.err
#SBATCH --partition=broadwl
#SBATCH --mem=36G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


python makeeQTLEmpirical_secondtry.py "Total" "/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/eQTL_total_EmpiricalDist_secondtry.txt"

python makeeQTLEmpirical_secondtry.py "Nuclear" "/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/eQTL_nuclear_EmpiricalDist_secondtry.txt"

QQ plots

qqplot(-log10(empiricalPval), -log10(actualPval))

Actual Dists:

nomnames=c("peakID", 'snp','dist', 'pval', 'slope')
eQTLinTotal=read.table("../data/eQTL_inAPA/eQTLinTotalApa_try2.txt", stringsAsFactors = F, col.names = nomnames)
eQTLinNuclear=read.table("../data/eQTL_inAPA/eQTLinNuclearApa_try2.txt", stringsAsFactors = F, col.names = nomnames)

Emiprical:

empTotal=read.table("../data/EmpiricalDists/eQTL_total_EmpiricalDist_secondtry.txt", col.names = nomnames,stringsAsFactors = F)
empNuclear=read.table("../data/EmpiricalDists/eQTL_nuclear_EmpiricalDist_secondtry.txt", col.names = nomnames, stringsAsFactors = F)

Deal with problem that there are less snps in empirical (add some uniform)

toaddTotal=runif(nrow(eQTLinTotal)-nrow(empTotal))
toaddNuclear=runif(nrow(eQTLinNuclear)-nrow(empNuclear))

Add the extra pvalues:

empNuclearUse= c(as.vector(empNuclear$pval),toaddNuclear)

empTotalUse= c(as.vector(empTotal$pval),toaddTotal)

make total plot:

png("../output/plots/eqtlsinTotAPAQQPlot_ALLeqtl.png")
qqplot(-log10(empTotalUse), -log10(eQTLinTotal$pval),ylab="-log10 Total APA pval", xlab="Empirical expectation", main="eQTLs in totalAPA analysis")
abline(0,1)
dev.off()

quartz_off_screen 
                2 

make nuclear:

png("../output/plots/eqtlsinNucAPAQQPlot_ALLeqtl.png")
qqplot(-log10(empNuclearUse), -log10(eQTLinTotal$pval),ylab="-log10 Nuclear APA pval", xlab="Empirical expectation", main="eQTLs in nuclearAPA analysis")

abline(0,1)
dev.off()

quartz_off_screen 
                2 

I will want to subset the actual to just get the unexplained ones:

/project2/gilad/briana/threeprimeseq/data/eQTL_Lietal/unexplained_FDR10.sort.txt

This has the unexplained eQTLs, I need to get all of the genes and switch symbol to gene name.

fixUnexplainedNames.py

  
geneNames=open("/project2/gilad/briana/genome_anotation_data/ensemble_to_genename.txt","r")
name_dic={}
for i,ln in enumerate(geneNames):
    if i >0:
        id, name, source =ln.split("\t")
        name_dic[id]=name 
inFile=open("/project2/gilad/briana/threeprimeseq/data/eQTL_Lietal/unexplained_FDR10.sort.txt","r")
outFile=open("/project2/gilad/briana/threeprimeseq/data/eQTL_Lietal/unexplained_FDR10_GENEnames.txt", "w")
nNot=0
for ln in inFile:
    line_list=ln.split()
    if len(line_list)==3:
        geneid=ln.split()[2]
        if geneid in name_dic.keys():
            genename=name_dic[geneid]
            outFile.write("%s\n"%(genename))
        else:
            nNot+=1
    else:
        geneList=line_list[2:]
        for each in geneList:
            if each in name_dic.keys():
                genename=name_dic[each]
                outFile.write("%s\n"%(genename))
            else:
                nNot+=1
print(nNot)
outFile.close()
          
          
          

Subset eQTLs for these unexplained eQTL gene.

subseteqtlUnexp.py

eQTLS_intot="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inTot.txt"
Unexp_eQTLS_intot="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_Unexplained_eQTLs_GeneNames_inTot.txt"
eQTLS_innuc="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_eQTLs_GeneNames_inNuc.txt"
Unexp_eQTLS_innuc="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_Unexplained_eQTLs_GeneNames_inNuc.txt"


def subsetQTL(inQTL, unexQTL):
    ungenes={}
    for ln in open("/project2/gilad/briana/threeprimeseq/data/eQTL_Lietal/unexplained_FDR10_GENEnames.txt","r"):
        gene=ln.strip()
        ungenes[gene]=""
    fout=open(unexQTL, "w")   
    for ln in open(inQTL, "r"):
        egene=ln.split()[0]
        if egene in ungenes.keys():
            fout.write(ln)
    fout.close()
    
subsetQTL(eQTLS_intot,Unexp_eQTLS_intot)
subsetQTL(eQTLS_innuc,Unexp_eQTLS_innuc)

Get read and empirical for these:

getApaPval4eQTLs_secondtry_UNEXP.py

def main(apa, eqtl, outFile):
    fout=open(outFile,"w")
    use_dic={}
    for ln in open(eQTL,"r"):
        egene=ln.split()[0]
        eSNP=ln.split()[5]
        use_dic[egene]=eSNP
    for ln in open(apa, "r"):
        geneApa=ln.split()[0].split(":")[-1].split("_")[0]
        snpApa =ln.split()[1]   
        if geneApa in use_dic.keys():
            if snpApa == use_dic[geneApa]:
                fout.write(ln)
    fout.close()
                

if __name__ == "__main__":
    import sys
    import random
    fraction=sys.argv[1]
    OutFile=sys.argv[2]
    if fraction=="Total":
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Total.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_Unexplained_eQTLs_GeneNames_inTot.txt"
    else:
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Nuclear.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_Unexplained_eQTLs_GeneNames_inNuc.txt"
    main(nomFile, eQTL, OutFile) 
    
    

run_getApaPval4eQTLs_second_unexp.sh

#!/bin/bash

#SBATCH --job-name=run_getApaPval4eQTLs2_unex
#SBATCH --account=pi-yangili1
#SBATCH --time=36:00:00
#SBATCH --output=run_getApaPval4eQTLs2_unex.out
#SBATCH --error=run_getApaPval4eQTLs2_unex.err
#SBATCH --partition=broadwl
#SBATCH --mem=36G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


python getApaPval4eQTLs_secondtry_UNEXP.py "Total" "/project2/gilad/briana/threeprimeseq/data/eQTL_inAPA/UnexpeQTLinTotalApa_try2.txt"

python getApaPval4eQTLs_secondtry_UNEXP.py "Nuclear" "/project2/gilad/briana/threeprimeseq/data/eQTL_inAPA/UnexpeQTLinNuclearApa_try2.txt"

makeeQTLEmpirical_secondtry_unex.py

def main(peaks, APA,nonE,eqtl, outF):
    #dictionary of non egenes  
    nonEgenes={}
    for ln in open(nonE):
        gene=ln.split()[0]
        nonEgenes[gene]=""
        
    # make dictionary- key= npeak, value=list of genes with same value (use this when i need to pull random from the values)
    #only add genes in the non eqet
    peaknum={}
    for ln in open(peaks, "r"):
        gene, npeak = ln.split()
        if npeak not in peaknum.keys():
            if gene in nonEgenes.keys():
                peaknum[npeak]=[gene]
        else:
            if gene in nonEgenes.keys():
                 peaknum[npeak].append(gene)
    #make opposite dic- key =gene, value = npeaks (use this when i first get an EGENE to see how many peaks i will need)
    #want all genes- e and no e genes 
    GenePeak={}
    for ln in open(peaks, "r"):
        gene, npeak= ln.split()
        GenePeak[gene]=npeak
    eQTLs= open(eqtl, "r")
    gene_useDic={}
    for ln in eQTLs:
        gene= ln.split()[0]
        peaks2use=GenePeak[gene]
        posibleGenes=peaknum[peaks2use]
        if len(posibleGenes) == 0:
             print("No genes with peaknum")
             posibleGenes=peaknum[str(int(peaks2use)+1)]
             if len(posibleGenes) == 0:
                 print("No genes with peaknum")
                 continue
        GeneUse=random.choice(posibleGenes)
        gene_useDic[GeneUse]=""
        #print(GeneUse)
        #find the snp that goes with this gene in the non egenes file- this gives me the lowest pval snp
    for ln in open(nonE,"r"):
        genenon=ln.split()[0]
        #print(genenon)
        if genenon in gene_useDic.keys():
            snpUse=ln.split()[5]
            gene_useDic[genenon]=snpUse
        #now i have the gene and the snp, i can write out any of the nominal values from the apa file with these
    fout=open(outF,"w")
    for ln in open(APA, "r"):
        geneApa=ln.split()[0].split(":")[-1].split("_")[0]
        snpApa =ln.split()[1]
        if geneApa in gene_useDic.keys():
            if snpApa == gene_useDic[geneApa]:
                fout.write(ln)
    fout.close()
  
  

if __name__ == "__main__":
    import sys
    import random
    fraction=sys.argv[1]
    OutFile=sys.argv[2]
    if fraction=="Total":
        npeaks="/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/TotalQTL_nPeaks.txt"
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Total.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        nonEfile="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inTot.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_Unexplained_eQTLs_GeneNames_inTot.txt"
    else:
        npeaks="/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/NuclearQTL_nPeaks.txt"
        nomFile="/project2/gilad/briana/threeprimeseq/data/nominal_APAqtl_GeneLocAnno_noMP_5percUs/filtered_APApeaks_merged_allchrom_refseqGenes.GeneLocAnno_NoMP_sm_quant.Nuclear.fixed.pheno_5perc.fc.gz.qqnorm_allNomRes.txt"
        nonEfile="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_NOTeQTLs_GeneNames_inNuc.txt"
        eQTL="/project2/gilad/briana/threeprimeseq/data/eQTL_myanalysis/permRes_Unexplained_eQTLs_GeneNames_inNuc.txt"
    main(npeaks, nomFile, nonEfile, eQTL, OutFile) 

run_makeeQTLEmpirical_secondtry_unexp.sh

#!/bin/bash

#SBATCH --job-name=run_makeeQTLEmpirical2_unexp
#SBATCH --account=pi-yangili1
#SBATCH --time=36:00:00
#SBATCH --output=run_makeeQTLEmpirical2_unexp.out
#SBATCH --error=run_makeeQTLEmpirical2_unexp.err
#SBATCH --partition=broadwl
#SBATCH --mem=36G
#SBATCH --mail-type=END

module load Anaconda3
source activate three-prime-env


python makeeQTLEmpirical_secondtry_unex.py "Total" "/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/eQTL_total_EmpiricalDist_secondtry_UNEXP.txt"

python makeeQTLEmpirical_secondtry_unex.py "Nuclear" "/project2/gilad/briana/threeprimeseq/data/EmpiricalDists/eQTL_nuclear_EmpiricalDist_secondtry_UNEXP.txt"

points(sort(-log10(runif(nrow(t)))), sort(-log10(t$corrPval)), col= alpha(“Red”)) abline(0,1)

UneQTLinTotal=read.table("../data/eQTL_inAPA/UnexpeQTLinTotalApa_try2.txt", stringsAsFactors = F, col.names = nomnames)
UNeQTLinNuclear=read.table("../data/eQTL_inAPA/UnexpeQTLinNuclearApa_try2.txt", stringsAsFactors = F, col.names = nomnames)

Emiprical:

empTotalUn=read.table("../data/EmpiricalDists/eQTL_total_EmpiricalDist_secondtry_UNEXP.txt", col.names = nomnames,stringsAsFactors = F)
empNuclearUn=read.table("../data/EmpiricalDists/eQTL_nuclear_EmpiricalDist_secondtry_UNEXP.txt", col.names = nomnames, stringsAsFactors = F)

Deal with problem that there are less snps in empirical (add some uniform)

toaddTotalUn=runif(nrow(UneQTLinTotal)-nrow(empTotalUn))
toaddNuclearUn=runif(nrow(UNeQTLinNuclear)-nrow(empNuclearUn))

Add the extra pvalues:

empNuclearUseUN= c(as.vector(empNuclearUn$pval),toaddNuclearUn)

empTotalUseUN= c(as.vector(empTotalUn$pval),toaddTotalUn)

make total plot:

png("../output/plots/eqtlsinTotAPAQQPlot.png")
qqplot(-log10(empTotalUse), -log10(eQTLinTotal$pval),ylab="-log10 Total APA pval", xlab="Empirical expectation", main="eQTLs in totalAPA analysis")
points(sort(-log10(empTotalUseUN)), sort(-log10(UneQTLinTotal$pval)),col= alpha("Red"))
legend("topleft", legend=c("All eQTLs", "Unexplained eQTLs"),col=c("black", "red"), pch=16,bty = 'n')
abline(0,1)
dev.off()

quartz_off_screen 
                2 

Nuclear plot:

png("../output/plots/eqtlsinNucAPAQQPlot.png")
qqplot(-log10(empNuclearUse), -log10(eQTLinNuclear$pval),ylab="-log10 Nuclear APA pval", xlab="Empirical expectation", main="eQTLs in nuclearAPA analysis")
points(sort(-log10(empNuclearUseUN)), sort(-log10(UNeQTLinNuclear$pval)),col= alpha("Red"))
legend("topleft", legend=c("All eQTLs", "Unexplained eQTLs"),col=c("black", "red"), pch=16,bty = 'n')
abline(0,1)
dev.off()

quartz_off_screen 
                2 

Same analysis starting with pQTLs

Make file with sig pQTLs:

Run this interactviely in R

library(dplyr)

names=c('pid', 'nvar', 'shape1', 'shape2','dummy', 'sid', 'dist', 'npval' ,'slope', 'ppval', 'bpval' )
permRes=read.table("/project2/gilad/briana/threeprimeseq/data/pQTL_myanalysis/fastqtl_qqnorm_prot.fixed.perm.out",stringsAsFactors = F, col.names = names)

permRes$bh=p.adjust(permRes$bpval, method="fdr")


permRes_sig=permRes %>% filter(-log10(bh)>=1)
nonSig=permRes %>% filter(-log10(bh)<1)


write.table(permRes_sig, file="/project2/gilad/briana/threeprimeseq/data/pQTL_myanalysis/permRes_significantpQTLs.txt", col.names=F, row.names=F, quote=F, sep="\t")

write.table(nonSig, file="/project2/gilad/briana/threeprimeseq/data/pQTL_myanalysis/permRes_NOTpQTLs.txt", col.names=F, row.names=F, quote=F, sep="\t")

Session information

sessionInfo()

R version 3.5.1 (2018-07-02)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS  10.14.1

Matrix products: default
BLAS: /Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRblas.0.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.5/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] cowplot_0.9.4   reshape2_1.4.3  forcats_0.4.0   stringr_1.4.0  
 [5] dplyr_0.8.0.1   purrr_0.3.1     readr_1.3.1     tidyr_0.8.3    
 [9] tibble_2.0.1    ggplot2_3.1.0   tidyverse_1.2.1 workflowr_1.2.0

loaded via a namespace (and not attached):
 [1] Rcpp_1.0.0       cellranger_1.1.0 plyr_1.8.4       pillar_1.3.1    
 [5] compiler_3.5.1   git2r_0.24.0     tools_3.5.1      digest_0.6.18   
 [9] lubridate_1.7.4  jsonlite_1.6     evaluate_0.13    nlme_3.1-137    
[13] gtable_0.2.0     lattice_0.20-38  pkgconfig_2.0.2  rlang_0.3.1     
[17] cli_1.0.1        rstudioapi_0.9.0 yaml_2.2.0       haven_2.1.0     
[21] xfun_0.5         withr_2.1.2      xml2_1.2.0       httr_1.4.0      
[25] knitr_1.21       hms_0.4.2        generics_0.0.2   fs_1.2.6        
[29] rprojroot_1.3-2  grid_3.5.1       tidyselect_0.2.5 glue_1.3.0      
[33] R6_2.4.0         readxl_1.3.0     rmarkdown_1.11   modelr_0.1.4    
[37] magrittr_1.5     whisker_0.3-2    backports_1.1.3  scales_1.0.0    
[41] htmltools_0.3.6  rvest_0.3.2      assertthat_0.2.0 colorspace_1.4-0
[45] stringi_1.3.1    lazyeval_0.2.1   munsell_0.5.0    broom_0.5.1     
[49] crayon_1.3.4