Kun:LabNotes/MONOD/2014-6-16
Jump to navigation
Jump to search
Analysis of the 140613_MiSeq data set[edit]
- Noi performed capture of 15 samples (Serum/tumor DNA from Pancreatic cancer patients, plasma from normal controls) using the MONOD V1N3 probe set:Noi/NOTES/2014-6-10
- A MiSeq run (151+6+151bp) was performed on this batch of libraries. Data folder: /home/kunzhang/seqStore/20140613_MiSeq
Low-level analysis[edit]
- A sample info sheet and a batch processing script was generated.
- All 15 data sets were mapped and the haploInfo files were generated.
./MONOD_batch_processing.pl 140613_MiSeq_sampleInfo.txt > MONOD_batch_processing.log & ../get_probeEfficiency_matrix.pl > 140613_MiSeq_probeEfficiency_matrix.txt
Sample_id | # raw reads | # mapped reads | % mapping | # on-target reads | % on-target | # UMIs | % clonal |
NC-P-10ng_1 | 1,651,384 | 821,536 | 49.7% | 703,453 | 85.6% | 47,641 | 93.23% |
NC-P-10ng_2 | 2,041,882 | 721,140 | 35.3% | 487,174 | 67.6% | 40,520 | 91.68% |
NC-P-5ng_1 | 2,126,484 | 653,912 | 30.8% | 422,917 | 64.7% | 22,424 | 94.70% |
NC-P-5ng_2 | 1,503,204 | 583,432 | 38.8% | 456,431 | 78.2% | 24,643 | 94.60% |
PC-S_10 | 1,998,558 | 920,290 | 46.0% | 542,561 | 59.0% | 31,570 | 94.18% |
PC-S_2 | 2,084,954 | 1,000,838 | 48.0% | 658,032 | 65.7% | 49,312 | 92.51% |
PC-S_3 | 1,780,602 | 878,368 | 49.3% | 661,506 | 75.3% | 36,352 | 94.50% |
PC-S_6 | 1,788,488 | 800,280 | 44.7% | 545,346 | 68.1% | 25,537 | 95.32% |
PC-S_7 | 2,042,144 | 578,860 | 28.3% | 194,299 | 33.6% | 12,612 | 93.51% |
PC-S_9 | 2,106,932 | 977,652 | 46.4% | 700,923 | 71.7% | 33,147 | 95.27% |
PC-T-1 | 2,643,028 | 1,773,208 | 67.1% | 1,463,016 | 82.5% | 105,276 | 92.80% |
PC-T-2 | 2,036,388 | 1,328,234 | 65.2% | 1,089,150 | 82.0% | 68,969 | 93.67% |
PC-T-4 | 2,632,888 | 1,567,146 | 59.5% | 1,149,534 | 73.4% | 83,254 | 92.76% |
PC-T-6 | 1,073,678 | 647,684 | 60.3% | 477,805 | 73.8% | 22,304 | 95.33% |
PC-T-7 | 2,646,776 | 1,618,226 | 61.1% | 1,175,527 | 72.6% | 86,816 | 92.61% |
allBED2MatrixFast.pl . > 140613_MiSeq_methylMatrix.txt #perform hierarchical clustering in R (need to manually clean up the sample names). x=read.table("140613_MiSeq_methylMatrix.txt",header=TRUE,row.names=1); x.cor = cor(x,use="pairwise.complete.obs",method="pearson") x.d = as.dist(1-x.cor) x.h = hclust(x.d,"complete") plot(x.h) File:140610 15 samples clustering.png
- Clearly healthy plasma and tumor samples are quite different, whereas patient serum cluster more closely to healthy plasma. So the real question is whether we can identify tumor-specific haplotypes from patient serum.