Dinh 2011/NOTES/2011-11-9
Jump to navigation
Jump to search
Localization of tested CpGs
File:AberrantMethylationLocalization 11102011.txt
- For files, see /media/1TB_store1/Dinh_Scratch/meangenemachine_desktop/MemoryMutations/DD_CpGLocalization/
File:NonOverlapping.hg18.wgEncodeRegDnaseClustered.txt
- This track shows DNase hypersensitive areas assayed in a large collection of cell types. Regulatory regions in general, and promoters in particular, tend to be DNase sensitive. Additional views of this dataset and additional documentation on the methods used for this track are available at the UW DNaseI HS page. The Peaks view in that page is the basis for the clusters shown here, which combine data from the peaks of the different cell lines in that page.
File:NonOverlapping.hg18.wgEncodeRegTfbsClustered.txt
- This track shows regions where transcription factors, proteins responsible for modulating gene transcription, bind to DNA as assayed by ChIP-seq (chromatin immunoprecipitation with antibodies specific to the transcription factor followed by sequencing of the precipitated DNA). Additional views of this dataset and additional documentation on the methods used for this track are available at the Yale TFBS Track and the HAIB TFBS Track pages. The Peaks views in those pages are the basis for the clusters shown here, which combine data from the peaks from the different cell lines and different transcription factors in those pages.
File:NonOverlapping.hg18.introns.txt
- Known intronic regions of any RefSeq gene.
File:NonOverlapping.hg18.exons.txt
- Known exonic regions of any RefSeq gene
File:NonOverlapping.hg18.3utr.txt
- Known 3UTR region of any RefSeq gene
File:NonOverlapping.hg18.5utr.txt
- Known 5UTR region of any RefSeq gene
File:NonOverlapping.hg18.promoter.txt
- Upstream 2kbp of any RefSeq TSS
- background counts
Gene | Unique_Tested_CpGs | 3utr | 5utr | exons | introns | promoters | sno_miRNA | wgEncodeRegDnaseClustered | wgEncodeRegTfbsClustered |
ASThiPS4F4 | 147647 | 7609 | 23432 | 29750 | 62694 | 27728 | 49 | 87019 | 75825 |
ASThiPS4F5 | 140219 | 7224 | 22197 | 28335 | 59422 | 26258 | 40 | 82799 | 71953 |
FiPS3F1 | 110438 | 5798 | 16883 | 21697 | 47101 | 19938 | 18 | 64078 | 55045 |
FiPS4F7 | 115079 | 6018 | 17716 | 22608 | 49201 | 20846 | 16 | 66658 | 57352 |
HUVhiPS4F1 | 142091 | 7276 | 22540 | 28415 | 60579 | 26487 | 24 | 83113 | 72206 |
HUVhiPS4F3 | 140528 | 7199 | 22172 | 27853 | 60127 | 26072 | 24 | 81813 | 70955 |
KhiPS4F8 | 118235 | 6151 | 18241 | 23017 | 50831 | 21428 | 18 | 68251 | 58857 |
MSCiPS4 | 192061 | 9735 | 30805 | 38516 | 81859 | 37187 | 55 | 112958 | 98686 |
MSCiPS8 | 195184 | 9867 | 31317 | 39052 | 83194 | 37739 | 55 | 114712 | 100162 |
PGP1-iPS-Repeat1 | 177496 | 9153 | 28132 | 35196 | 76380 | 33917 | 44 | 102974 | 89690 |
- background rates are similar overall
Gene | Unique_Tested_CpGs | 3utr | 5utr | exons | introns | promoters | sno_miRNA | wgEncodeRegDnaseClustered | wgEncodeRegTfbsClustered |
ASThiPS4F4 | 147647 | 5% | 16% | 20% | 42% | 19% | 0% | 59% | 51% |
ASThiPS4F5 | 140219 | 5% | 16% | 20% | 42% | 19% | 0% | 59% | 51% |
FiPS3F1 | 110438 | 5% | 15% | 20% | 43% | 18% | 0% | 58% | 50% |
FiPS4F7 | 115079 | 5% | 15% | 20% | 43% | 18% | 0% | 58% | 50% |
HUVhiPS4F1 | 142091 | 5% | 16% | 20% | 43% | 19% | 0% | 58% | 51% |
HUVhiPS4F3 | 140528 | 5% | 16% | 20% | 43% | 19% | 0% | 58% | 50% |
KhiPS4F8 | 118235 | 5% | 15% | 19% | 43% | 18% | 0% | 58% | 50% |
MSCiPS4 | 192061 | 5% | 16% | 20% | 43% | 19% | 0% | 59% | 51% |
MSCiPS8 | 195184 | 5% | 16% | 20% | 43% | 19% | 0% | 59% | 51% |
PGP1-iPS-Repeat1 | 177496 | 5% | 16% | 20% | 43% | 19% | 0% | 58% | 51% |
File:SharedAberrantCpGsLocalization.xls
- 9% of genes have greater rates than the background of CpGs in 3UTR
- 30% of genes have greater rates than the background of CpGs in 5UTR
- 24% of genes have greater rates than the background of CpGs in exons
- 81% of genes have greater rates than the background of CpGs in introns
- 7% of genes have greater rates than the background of CpGs in promoter (2kp upstream tss)
Since these CpGs were selected to be around genes, we already expect to find sites in gene bodies; there is a clear bias toward intronic regions, which is not seen in the background localization of CpGs.
- 45% of genes have greater rates than the background of CpGs in Dnase I clusters
- 33% of genes have greater rates than the background of CpGs in TF binding sequence
- At first glance, shared aberrant CpGs tend to not be in coding regions and are more likely to be in introns or DNaseI hypersensitivity and TF binding sequence.
- If shared aberrant CpGs are in regulatory regions then how do they regulate the expression of these genes? Is there a correlation? If it is a weak correlation, could it be possible that multiple regulatory "switches" act on this gene to control it's transcription and degradation? Can we find those?
After discussion with Athurva
- I am not getting the same number of testable sites for each iPSCs as Athurva. But I can just get a list from Athurva later.
- I need to re-run the script because I missed NSCiPS line
- Also, for the background, we should only consider CpG within a certain window of those shared genes
- The File:SharedAberrantCpGsLocalization.xls shows that parsing of the RefSeq gene file was wrong - for example, some sites fall on both 3UTR and 5UTR
- 1) There are genes which are in 3UTR as well as 5UTR: this is because I forgot to make sure that noncoding genes have no 3UTR and no 5UTR. RefSeq annotated these non protein coding genes with cdsStart and cdsEnds which are at the same position
- 2) Not all genes have both 3UTR and 5UTR: the script I used made the entire transcript as 3UTR/5UTR when there should really be none.
- 3) It could be interesting to separate out protein coding and non-protein mRNA in refSeq so that we can better separate out methylation which is regulating protein structure with alternative splicing or regulating the transcription/degradation of mRNAs.