Kun:LabNotes/MONOD/2015-7-16

From ZhangLabWiki
Jump to navigation Jump to search

MONOD V6 (aka MONOD June2015)[edit]

Probe Design[edit]

  • Target selection:
   1. MONOD v1/v3 efficient targets
   2. BluePrint efficient targets
   3. Illumina/USC 1000 CRC targets
   4. Hand curated colon cancer methylation markers
   5. Hypermethylated genes from Hyen & Esteller NRG 2012 (Table 1&3)
   6. Commonly methylated genes from Ciriello et al NG 2013 (Figure 1C)
   7. Guo TCGA top 75 sites
   8. Guo TCGA NSCLC 5 sites
   9. Dana's gene list (BRCA1, MLH and related genes)
  • Design log: (Genome-miner: /home/kunzhang/CpgMIP/MONOD/June2015_probe_design)
   sort -k 1,1 -k 2,2n MONOD_June2015_all_targets_bed.txt | awk '{print $1"\t"$2"\t"$3}' | /home/kunzhang/softwares/bedtools-2.20/bin/bedtools merge -d 40 -i - > MONOD_June2015_all_targets_merged_bed.txt
   awk '{print $1":"$2"w\t"$1"\t"$2"\t"$3"\t+"'} MONOD_June2015_all_targets_merged_bed.txt > MONOD_June2015_all_targets_merged_ppDesigner.in
   awk '{print $1":"$2"c\t"$1"\t"$2"\t"$3"\t-"'} MONOD_June2015_all_targets_merged_bed.txt >> MONOD_June2015_all_targets_merged_ppDesigner.in
   split -l 1000 MONOD_June2015_all_targets_merged_ppDesigner.in MONOD_June2015_all_targets_merged_ppDesigner.in.
   sed 's/aa/ab/' < jobFile_June2015.pl > jobFile_June2015_ab.pl
   sed 's/aa/ac/' < jobFile_June2015.pl > jobFile_June2015_ac.pl
   sed 's/aa/ad/' < jobFile_June2015.pl > jobFile_June2015_ad.pl
   sed 's/aa/ae/' < jobFile_June2015.pl > jobFile_June2015_ae.pl
   sed 's/aa/af/' < jobFile_June2015.pl > jobFile_June2015_af.pl
   sed 's/aa/ag/' < jobFile_June2015.pl > jobFile_June2015_ag.pl
   sed 's/aa/ah/' < jobFile_June2015.pl > jobFile_June2015_ah.pl
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015.pl > June2015_probes.aa.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_ab.pl > June2015_probes.ab.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_ac.pl > June2015_probes.ac.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_ad.pl > June2015_probes.ad.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_ae.pl > June2015_probes.ae.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_af.pl > June2015_probes.af.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_ag.pl > June2015_probes.ag.txt &
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl jobFile_June2015_ah.pl > June2015_probes.ah.txt &
   cat June*.txt > MONOD_June2015_all_targets_probeInfo.txt
   grep probes MONOD_June2015_all_targets_probeInfo.txt | awk '{print $1}' > MONOD_June2015_1st_design_failed_targets.txt
   match_column.pl MONOD_June2015_1st_design_failed_targets.txt MONOD_June2015_all_targets_merged_ppDesigner.in 1 > MONOD_June2015_all_targets_merged_ppDesigner_2nd.in
   # allow for up to 2 CpG sites per arm
   /home/kunzhang/softwares/ppDesigner_Kun_version/src/ppDesigner.pl  jobFile_June2015_2nd.pl > June2015_probes_2nd.txt &

Experiments[edit]

First test[edit]

  • Dinh did the first test:Dinh/Dinh_2015/NOTES/2015-6-25.
  • The sequencing data are in genome-miner: /media/LTS_33T/SeqStore2/150721_MiSeq
  • I did the checking for target-independent circularization, by taking 100,
  #in the working directory: /home/kunzhang/CpgMIP/MONOD/Data/150721_MiSeq
  ./batch_map_N_parse.pl 
  ./get_small_insert_count_matrix.pl > 150721_MiSeq_small_insert_count_matrix.txt 
  • I then take the top 100 probes with the most target-independent circularization events, and did the clustering with R.
  x=read.table("150721_MiSeq_small_insert_count_matrix_top100.txt",header=TRUE,row.names=1);
  library(gplots)
  heatmap.2(as.matrix(log10(x+1)),scale="none", key=T, symkey=F,density.info="none",trace="none",dendrogram="both",Rowv=TRUE,Colv=TRUE, cexCol=0.6, cexRow=0.3)
   Hi-Res PDF
   File:150721 MiSeq small insert count top100 heatmap.png
  • Observations:
    • BSPP capture from gDNA or fragmented DNA is cleaner than from WGBS libraries.
    • The mean counts for BSPP and those for WGBS-BSPP are still correlated, especially after taking log transformation.
    • The top 12 probes represent 70% of the target-independent circularization events from gDNA, and 46% from WGBS libraries. Design suppressor oligos for these probes:
probe_id Mean(BSPP) H1 H1_Tm H2 H2_Tm Suppressor_H2
chr14:75516400c_2568 227.88 AGAGTTGAGTAAAGGTCGGGTG 62 TCGAGATTAGTTTGATTAATATGGAGAA 59 TCAGGATACACACTACCCTCGAGATTAGTTTGATTAATATGGAGAA
chr12:111618822w_195 146.00 AAATTATTTGATTGAAATAATTAGGCGG 58 TGTTAGAGGCGGTTTGTGTATT 60 TCAGGATACACACTACCCTGTTAGAGGCGGTTTGTGTATT
chr3:50352990c_471 36.38 AATTAGTCGGGTATGGTGGTGTTTGTTT 65 TGGAGGTTGTAGTGAGTTAAGA 59 TCAGGATACACACTACCCTGGAGGTTGTAGTGAGTTAAGA
chr14:75516400c_972 29.88 GGGAGGATTATTTGAGTTTAGGAGTT 61 AAATTAGTTAGGTATGGTGGCGAG 61 TCAGGATACACACTACCCAAATTAGTTAGGTATGGTGGCGAG
chr13:114909082c_105 28.38 TATTGTTATATGGTGCGTGGTTGTATTT 62 TGTTATGTGGTGCGTGATTGTA 61 TCAGGATACACACTACCCTGTTATGTGGTGCGTGATTGTA
chr2:48013200w_1241 21.50 TTTGGGAGATTGAGATGGGAGGAT 64 TTTAGTTTGGGTGGTAGAGTTAGATT 60 TCAGGATACACACTACCCTTTAGTTTGGGTGGTAGAGTTAGATT
chr20:32378088c_130 17.63 TTATTTATAGGATTATAGGTTAGGCGT 58 TTATTTGAGGTTAGGAGTTCGAT 58 TCAGGATACACACTACCCTTATTTGAGGTTAGGAGTTCGAT
chr3:50352990c_186 14.75 TTTAGTTTGGGTGATAGAGTAAGATTT 59 GAGAAAGGAGTTGTGGTTTGTTT 60 TCAGGATACACACTACCCGAGAAAGGAGTTGTGGTTTGTTT
chr6:137242190w_106 14.13 TGAATTTAGAATAAAGAGTTATCGAAGT 58 TTTGTGGTGTAAAGGTGTGAAT 59 TCAGGATACACACTACCCTTTGTGGTGTAAAGGTGTGAAT
chr5:132161518c_356 14.00 GTATTTATTTGCGTAGTTATGGAGGT 60 TGGATTTGTGGTATTTGTTTATGC 59 TCAGGATACACACTACCCTGGATTTGTGGTATTTGTTTATGC
chr7:6047900w_2500 13.75 TTATTATGTTGGTTAGGTTGATTTCG 58 ATTATAGGTGTGAGTTATCGTGTT 58 TCAGGATACACACTACCCATTATAGGTGTGAGTTATCGTGTT
chr14:105119324c_429 10.00 AGGTTTGGATATATTAGGTTCGTT 58 GTTTGGATATATTAGGTTCGTTTAGT 58 TCAGGATACACACTACCCGTTTGGATATATTAGGTTCGTTTAGT