# Imports
%matplotlib inline
%config InlineBackend.figure_format = 'retina'
from basepair.imports import *
hv.extension('bokeh')

Using TensorFlow backend.

from pathlib import Path

# Common paths
model_dir = Path(f"{ddir}/processed/chipnexus/exp/models/oct-sox-nanog-klf/models/n_dil_layers=9/")
modisco_dir = model_dir / f"modisco/all/deeplift/profile/"
output_dir = Path("/srv/www/kundaje/avsec/chipnexus/oct-sox-nanog-klf/models/n_dil_layers=9/modisco/all/deeplift/profile")

from basepair.modisco.pattern_instances import load_instances, filter_nonoverlapping_intervals, plot_coocurence_matrix

from basepair.exp.paper.config import motifs

dfi = load_instances(f"{modisco_dir}/instances.parq", motifs)

number of deduplicatd instances: 1594551 (35.14740608536545%)

dfi.head()

	pattern	example_idx	pattern_start	pattern_end	strand	pattern_len	pattern_center	match_weighted	match_weighted_p	match_weighted_cat	match_max	match_max_task	imp_weighted	imp_weighted_p	imp_weighted_cat	imp_max	imp_max_task	seq_match	seq_match_p	seq_match_cat	match/Klf4	match/Nanog	match/Oct4	match/Sox2	imp/Klf4	imp/Nanog	imp/Oct4	imp/Sox2	example_chrom	example_start	example_end	example_strand	example_interval_from_task	pattern_short	pattern_name	pattern_start_abs	pattern_end_abs
0	metacluster_0/pattern_0	0	104	119	+	15	111	0.2469	0.0010	low	0.4484	Oct4	0.0365	NaN	NaN	0.0524	Oct4	3.9825	0.0085	low	0.2654	-0.3593	0.4484	0.3725	0.0235	0.0246	0.0524	0.0313	chrX	143482544	143483544	*	Oct4	m0_p0	Oct4-Sox2	143482648	143482663
1	metacluster_0/pattern_0	0	263	278	-	15	270	0.2518	0.0011	low	0.2893	Oct4	0.0217	NaN	NaN	0.0342	Nanog	1.1066	0.0007	low	0.1618	0.2170	0.2893	0.2803	0.0143	0.0342	0.0200	0.0203	chrX	143482544	143483544	*	Oct4	m0_p0	Oct4-Sox2	143482807	143482822
2	metacluster_0/pattern_0	0	282	297	+	15	289	0.3466	0.0069	low	0.3870	Sox2	0.0272	NaN	NaN	0.0296	Sox2	5.0100	0.0188	low	0.2257	0.3454	0.3723	0.3870	0.0233	0.0258	0.0279	0.0296	chrX	143482544	143483544	*	Oct4	m0_p0	Oct4-Sox2	143482826	143482841
3	metacluster_0/pattern_0	0	445	460	+	15	452	0.2399	0.0008	low	0.2469	Sox2	0.1852	0.0008	high	0.4145	Nanog	1.7775	0.0012	low	0.2458	0.2408	0.2314	0.2469	0.1234	0.4145	0.0949	0.1896	chrX	143482544	143483544	*	Oct4	m0_p0	Oct4-Sox2	143482989	143483004
4	metacluster_0/pattern_0	0	475	490	-	15	482	0.2903	0.0025	low	0.3322	Oct4	0.2889	0.0179	high	0.5196	Nanog	3.3390	0.0057	low	0.2092	0.2681	0.3322	0.2991	0.1613	0.5196	0.2049	0.3241	chrX	143482544	143483544	*	Oct4	m0_p0	Oct4-Sox2	143483019	143483034

dfi.to_csv(modisco_dir / 'instances.motifs.tsv', sep='\t')

!cp {modisco_dir}/instances.motifs.tsv {output_dir}/instances.motifs.tsv

# create_tf_session(0)