In [1]:
# Parameters
data_name = "Ar"
task_idxs = "0-213,216-278"
homer_root = "/srv/scratch/msharmin/mouse_hem/with_tfd/full_mouse50/Naive_scans"
modisco_root = "/srv/scratch/msharmin/mouse_hem/with_tfd/full_mouse50/Naive_modisco2019"
overwrite = False
In [2]:
from matlas.reports import collect_motifDB_to_be_displayed, display_comparative_cell_types
load data from labcluster
Using TensorFlow backend.
2019-08-23 10:57:28,783 [WARNING] git-lfs not installed
In [3]:
homer_matches, modisco_matches = collect_motifDB_to_be_displayed(task_idxs, homer_root, modisco_root, 
                                                              overwrite=False)
TF-MoDISco is using the TensorFlow backend.
In [4]:
display_comparative_cell_types(data_name, homer_matches, modisco_matches)
Ar: Motif: M03436_2.00



Forward PWM

Reverse PWM

Shared Cell types both TF-MoDISco and Homer finds Ar: #1

Cell typeModiscoHomer
MTR_NRN_MN1;methyl3;D;En

Pattern: metacluster_0/pattern_10

Significance: 0.030849599999999998

Pattern: motif8.motif

Significance: 0.030648599999999998

Unique Cell types only TF-MoDISco finds Ar: #3

Cell typeModiscoHomer
Adipose;D;En

Pattern: metacluster_0/pattern_18

Significance: 0.00530268

/

Pattern: metacluster_0/pattern_19

Significance: 0.019643599999999997

Absent
SKLTL_Muscle;8wks;D;En

Pattern: metacluster_1/pattern_18

Significance: 0.00541172

Absent
Muscle;A;CL

Pattern: metacluster_1/pattern_14

Significance: 0.011103

Absent
No Cell types only Homer finds Ar