Data-driven selection of cell line models and epigenetic comparison across cell types
-
18/07/2017
-
Working on ATAC data
-
Current matrix lists
- As below, have added gsk_chip_filtered, all GSK data so far (with ATAC-seq added), size = 4 x 43 x 444967
-
Looking at MDS distances for Projects 1-3.
-
-
12/07/2017
-
Current matrix lists
-
all_data - Blueprint + Project 1 data + some lung cell line outliers, size = 50 x 18068, generated from vignettes/project_1.R, TSSs of protein-coding genes
-
mask_data - Blueprint + Project 1/3 + ENCODE, size = 112 x 444967, generated from test/test_masking.R, all regulatory regions (defined by Ensembl multi-cell build)
-
mask_data (Aidan, local) - As above with Project 2 data added, size = 137 x 444967
-
-
Current status
-
Graeme (Tessella) - working with mask_data using a combination of PCA and PLS (?) to attempt to uncover an immortalisation signature
-
Nam - working with all_data using hierarchical clustering on combined data types to model the haematopoietic lineage hieraerchy
-
David - Using gene/datatype clustering to look for differential patterns between cell groups
-
-