Noi/NOTES/2012-2-15

From ZhangLabWiki
Jump to navigation Jump to search

Link to calendar:[[1]]

  • Continued from 2012_02_02: [[2]]

ASM analysis on the two pedigrees of HAPMAP PT01 samples[edit]

Samples information[edit]

  • HAPMAPPT01: [3]
Catalog ID Sex Family Relationship
GM11995 Female 1362 maternal grandmother
GM11993 Female 1362 paternal grandmother
GM10861 Female 1362 mother
GM11992 Male 1362 paternal grandfather
GM11994 Male 1362 maternal grandfather
GM10860 Male 1362 father
GM12813 Female 1454 paternal grandmother
GM12802 Female 1454 mother
GM12815 Female 1454 maternal grandmother
GM12812 Male 1454 paternal grandfather
GM12814 Male 1454 maternal grandfather
GM12801 Male 1454 father


  • Retrieved the genotypes of HAPMAP samples from,
http://hapmap.ncbi.nlm.nih.gov/downloads/genotypes/2008-07_phaseIII/
http://hapmap.ncbi.nlm.nih.gov/downloads/genotypes/2008-07_phaseIII/hapmap_format/forward/
  • Note: I didn't download from the latest version since it didn't include some samples I am analyzing.
  • GO.wget.sh: Media:GO.wget.txt
  • Concatenate genotypes in from every chromosome into the same file: HAPMAP.genotypes_CEU_phase3.1_nr.b36_fwd.txt
  • Print out only the column containing the genotypes of samples in family 1362 and 1454:
to list the number of column: 
head -1 HAPMAP.genotypes_CEU_phase3.1_nr.b36_fwd.txt | sed s'/\ /\n/g' > tmp
awk '{print $1" "$2" "$3" "$4" "$5" "$76" "$75" "$78" "$77" "$74" "$79" "$138" "$142" "$137" "$140" "$139" "$141;}' HAPMAP.genotypes_CEU_phase3.1_nr.b36_fwd.txt > HAPMAP1362-1454.genotypes_CEU_phase3.1.txt
1,402,288 sites
  • Since the HAPMAP genotype files were built in hg18, I used Dinh's script to lift over snp's RS values to hg19 positions.
  • liftoverRSvalues.pl: Media:liftoverRSvalues.txt
./liftoverRSvalues.pl /GenomeDB/HsGenome/snp134_snv.txt <HAPMAP1362-1454.genotypes_CEU_phase3.1.txt>HAPMAP1362-1454.genotypes_phase3.1.hg19snp134
    • Note: after using lift over, I found the numbers of SNPs before and after lift over were different by increasing more than 30,000 sites. I would figure out what is the cause.
wc -l HAPMAP_1363-1454_genotypes.txt --> 1,402,289 SNPs
wc -l HAPMAP_1363-1454_genotypes_hg19.snp134 --> 1,438,444 SNPs
1438444 - 1,402,288 =  36,156 SNPs
  • Dinh suggested to remove the sites in column3 containing 6_apd_hap1 or *_random etc..
awk '{if ($3 !~ /_/) print $0;}' HAPMAP1362-1454.genotypes_phase3.1.hg19snp134 > HAPMAP1362-1454.genotypes_phase3.1.hg19snp134.txt
  • I got total 1,400,104 sites

On ASM output data[edit]

  • Since MFASMAnalysisQValues.GM*.cpg.txt files contain a list of SNP coordinates, SNP called (allele1 and allele2), every single CpG site with chr_pos also the number of C's and T's associated with SNP base1 and base2, I chose to work on these files and cp them in the same foleder: /home/nplongth/Noi_scratch/ASM_HAPMAP1362-1454_2012_02_03/MFASMAnalysisQValues-cpg
ls MFASMAnalysisQValues-cpg/MFASMAnalysisQValues.GM*.cpg.txt > tmp
awk '{print "123"$1"123\t"$1;}' tmp | sed s'/123MFASMAnalysisQValues-cpg\/MFASMAnalysisQValues.GM/NA/g' | sed s'/1362.cpg.txt123//g' | sed s'/1454.cpg.txt123//g'> MFASM.QValues-cpg.list.txt --> 74,555 sites
  • Then I sorted out the matrix based on the number of methylation value available in each CpG site by excel. Then get the new matrix with the minimum number of methylation values 4. There are total 20,246 CpG sites left.
  • Calculate ...
./calSNPCpG_MFASM.QValues-cpg.pl sorted_HAPMAP.SNPCpG_MFASM.QValues-cpg_min4values.txt > calDist_HAPMAP.SNPCpG.QValues-cpg_min4values.txt
  • Print out only the sites show sequence-dependent (candidate sites)
awk '{if ($0 ~ /SEQ_DEP/) print $0;}' calDist_HAPMAP.SNPCpG.QValues-cpg_min4values.txt  > SEQ_DEP.HAPMAP.SNPCpG.QValues-cpg_min4values.txt (4,862 sites)
  • Currently, I couldn't find a good cutoff to identify the significant ASM.
  • Continued on 2012_02_27: [[4]]