ATAC-Seq: repetitive elements analysis

[dktanwar]

Hierarchy of RE

LTR (Family) —> L1 (Sub-Family) —> L1_Rod (Gene) —> L1_Rod_dup1 (Transcript)
LTR (Family) —> L1 (Sub-Family) —> L1_Rod (Gene) —> L1_Rod_dup2 (Transcript)
LTR (Family) —> L1 (Sub-Family) —> L1_Rod (Gene) —> L1_Rod_dup3 (Transcript)

In our analysis, we refer to Gene as subcategory and Transcript as locus.

Differential accessibility analysis of repetitive elements

Directory: 20200129-01_repetetive_elements_atac_diff_acc

• [x] DA at the locus level
• [x] DA at the subcategory level
• [x] Overlap ATAC peaks with locus and subcategory, a minimum overlap of 5bp
• [x] Overlap ChIP peaks with locus and subcategory, a minimum overlap of 5bp

Transcription factor motif analysis and GO analysis

1. For each of the differentially accessible subcategory, select the differentially accessible loci.
2. Remove the subcategories which have less than 10 differentially accessible loci.

Categorization of input:

1. All the RE together
2. RE which has more accessibility in adult
3. RE which has less accessibility in adult

Data analysis

[dktanwar]

From #40

We have performed HOMER using the class as a background (eg: all LTRs, all SINEs). It would be good to perform GREAT similarly (now we either have run GREAT on all identified REs or on all diff access REs). I would need:

• [x] GREAT analysis on Categories 1-6 using class as a background (the excel sheet with the categories you should already have, but if not let me now and I can resend it).

Eg: Category 1 (less access ERV1 subtypes ran against all identified LTRs); Category 2 (less access ERVK subtypes ran against all identified LTRs); Category 5 (more access ERVK subtypes ran against all LTRs); Category 6 (more access L1 subtypes ran against all identified LINEs) etc....

[dktanwar]

Less accessible in Adult

• GO_BP: Category 3 have significant results

More accessible in Adult

• GO_BP: Category 4 have significant results
• GO_BP: Category 6 have significant results