Wiki
Add details about figures to wiki (add seperate figures page)
Add column to metadata
- paper annotations
Papers to read
Pu.1.pdf in teams is highest priority. (only looks at one cell type).
DONE Figure 1 (UMAP)
Make UMAP instead of PCA.
- What is the original dimension of the space?
Expression data
PU1 peak count doesn't match expression.
- Protein Atlas SPI1 to get PU.1 (6)
- Casey might know about Databases containing single cell expression data
Figure 2
- GREAT has a peak limit so we can't find nearby genes.
- nearest gene?
- Bring Figure 2 to lab meeting
B.
Remake figure 2.B Shreeja has code.
DONE C. (Pathway Analysis)
- Hallmark Pathway analysis (bring up in computational meeting).
-
ANSWER
We should subsample peaks for GREAT
Figure 3
Order columns by amount of SPI1 (ets) and move SPI1 to bottom.
Change long (small font) name to Shared PU.1 Peaks
B.
- SPAMO: MEME suite. Gives related motifs (co-occuring). html reports are in project directory.
- look at other papers to see spamo results.
C.
Genome browser shots. PU.1 related motifs from literature
Figure 4 Mario (Top Priority)
2018 John Harley, Matt (MARIO)
Mario looks at allelic imbalance. Should be run on as many cell lines as possible (Bullock and neutrophils).
Requires genotype data.
- Download VCF files from EGA (EGAC00001000135)
Pathway analysis
How should we do it?
Phenotype Reli (low priority)
- Need to wait
Extra notes
Look up Wilbur Lam MD, PhD
Anemia photo detector (fingernails).