from collections import OrderedDict
exp = 'nexus,peaks,OSNK,0,10,1,FALSE,same,0.5,64,25,0.004,9,FALSE,[1,50],TRUE'
imp_score = 'profile/wn'

motifs = OrderedDict([
    ("Oct4-Sox2", 'Oct4/m0_p0'),
    ("Oct4", 'Oct4/m0_p1'),
    # ("Strange-sym-motif", 'Oct4/m0_p5'),
    ("Sox2", 'Sox2/m0_p1'),
    ("Nanog", 'Nanog/m0_p1'),
    ("Zic3", 'Nanog/m0_p2'),
    ("Nanog-partner", 'Nanog/m0_p4'),
    ("Klf4", 'Klf4/m0_p0'),
])

# Imports
from basepair.imports import *
from plotnine import *
import plotnine
import warnings
warnings.filterwarnings("ignore")

%matplotlib inline
%config InlineBackend.figure_format = 'retina'
paper_config()

# interval columns in dfi
interval_cols = ['example_chrom', 'pattern_start_abs', 'pattern_end_abs']

Using TensorFlow backend.

from basepair.exp.paper.config import *

# figures dir
model_dir = models_dir / exp
fdir = Path(f'{ddir}/figures/modisco/{exp}/spacing/')
fdir_individual = fdir / 'individual'
fdir_individual_sim = fdir / 'individual-simulation'

dfab = pd.read_csv(f"{model_dir}/deeplift/dfab.csv.gz")

dfab.head()

df = dfab[(dfab.center_diff <= 100) & 
          (dfab.motif_pair == 'Nanog<>Nanog')]

motif_pair_name = 'Nanog<>Nanog'
plotnine.options.figure_size = get_figsize(2, aspect=2/10*4 / 2)
xmin = 5
xmax = 100
fig = (ggplot(aes(x='center_diff', fill='strand_combination'), df) + 
 # plot
 geom_histogram(breaks=np.arange(xmin, xmax+1)) + facet_grid("strand_combination~.") + 
 # Theme, labels, colors
 theme_bw(base_size=10, base_family='Arial') + 
 theme(strip_text = element_text(rotation=0), legend_position='top') +  
 xlab("Pairwise distance") +
 ggtitle(motif_pair_name) + 
 scale_x_continuous(breaks=np.arange(xmin, xmax, step=5),
                    minor_breaks=np.arange(xmin, xmax, step=1)) +
 scale_fill_brewer(type='qual', palette=3))
 # axis_ticks_major_x()
display(fig)
# fig.save(fdir_individual / f'{motif_pair_name}.large.pdf')

<ggplot: (8755103742370)>

df.head()

	example_chrom_x	example_end_x	example_idx	example_interval_from_task_x	example_start_x	example_strand_x	imp/Klf4_x	imp/Nanog_x	imp/Oct4_x	imp/Sox2_x	imp_max_x	imp_max_task_x	imp_weighted_x	imp_weighted_cat_x	imp_weighted_p_x	match/Klf4_x	match/Nanog_x	match/Oct4_x	match/Sox2_x	match_max_x	match_max_task_x	match_weighted_x	match_weighted_cat_x	match_weighted_p_x	pattern_x	pattern_center_x	pattern_end_x	pattern_end_abs_x	pattern_len_x	pattern_name_x	pattern_short_x	pattern_start_x	pattern_start_abs_x	seq_match_x	seq_match_cat_x	seq_match_p_x	strand_x	tf_x	pattern_center_aln_x	pattern_strand_aln_x	row_idx_x	id_x	Klf4/profile_match_x	Klf4/profile_match_p_x	Klf4/profile_counts_x	Klf4/profile_counts_p_x	Klf4/profile_max_x	Klf4/profile_max_p_x	Klf4/profile_counts_max_ref_x	Klf4/profile_counts_max_ref_p_x	Nanog/profile_match_x	Nanog/profile_match_p_x	Nanog/profile_counts_x	Nanog/profile_counts_p_x	Nanog/profile_max_x	Nanog/profile_max_p_x	Nanog/profile_counts_max_ref_x	Nanog/profile_counts_max_ref_p_x	Oct4/profile_match_x	Oct4/profile_match_p_x	Oct4/profile_counts_x	Oct4/profile_counts_p_x	Oct4/profile_max_x	Oct4/profile_max_p_x	Oct4/profile_counts_max_ref_x	Oct4/profile_counts_max_ref_p_x	Sox2/profile_match_x	Sox2/profile_match_p_x	Sox2/profile_counts_x	Sox2/profile_counts_p_x	Sox2/profile_max_x	Sox2/profile_max_p_x	Sox2/profile_counts_max_ref_x	Sox2/profile_counts_max_ref_p_x	example_chrom_y	example_end_y	example_interval_from_task_y	example_start_y	example_strand_y	imp/Klf4_y	imp/Nanog_y	imp/Oct4_y	imp/Sox2_y	imp_max_y	imp_max_task_y	imp_weighted_y	imp_weighted_cat_y	imp_weighted_p_y	match/Klf4_y	match/Nanog_y	match/Oct4_y	match/Sox2_y	match_max_y	match_max_task_y	match_weighted_y	match_weighted_cat_y	match_weighted_p_y	pattern_y	pattern_center_y	pattern_end_y	pattern_end_abs_y	pattern_len_y	pattern_name_y	pattern_short_y	pattern_start_y	pattern_start_abs_y	seq_match_y	seq_match_cat_y	seq_match_p_y	strand_y	tf_y	pattern_center_aln_y	pattern_strand_aln_y	row_idx_y	id_y	Klf4/profile_match_y	Klf4/profile_match_p_y	Klf4/profile_counts_y	Klf4/profile_counts_p_y	Klf4/profile_max_y	Klf4/profile_max_p_y	Klf4/profile_counts_max_ref_y	Klf4/profile_counts_max_ref_p_y	Nanog/profile_match_y	Nanog/profile_match_p_y	Nanog/profile_counts_y	Nanog/profile_counts_p_y	Nanog/profile_max_y	Nanog/profile_max_p_y	Nanog/profile_counts_max_ref_y	Nanog/profile_counts_max_ref_p_y	Oct4/profile_match_y	Oct4/profile_match_p_y	Oct4/profile_counts_y	Oct4/profile_counts_p_y	Oct4/profile_max_y	Oct4/profile_max_p_y	Oct4/profile_counts_max_ref_y	Oct4/profile_counts_max_ref_p_y	Sox2/profile_match_y	Sox2/profile_match_p_y	Sox2/profile_counts_y	Sox2/profile_counts_p_y	Sox2/profile_max_y	Sox2/profile_max_p_y	Sox2/profile_counts_max_ref_y	Sox2/profile_counts_max_ref_p_y	center_diff	center_diff_aln	strand_combination	motif_pair	Nanog/profile_counts_max_ref_x_log	Nanog/profile_counts_max_ref_y_log	Sox2/profile_counts_max_ref_x_log
261310	chr3	1.2215e+08	1	Oct4	1.2215e+08	.	NaN	1.1791	NaN	NaN	1.1791	Nanog	1.1791	high	0.9987	NaN	0.5380	NaN	NaN	0.5380	Nanog	0.5380	medium	0.4364	Nanog/metacluster_0/p...	464.0	468.0	1.2215e+08	9.0	Nanog	Nanog/m0_p1	459.0	1.2215e+08	3.9726	low	0.1077	+	Nanog	464.0	-	263219.0	263219.0	5.1041	0.5008	57.0001	0.5666	4.0	0.7477	2.0000e-06	0.7316	3.9462	0.3717	3579.0000	0.9963	281.0	0.9987	1.0	0.6347	3.5280	0.2090	2202.0000	0.9993	104.0	0.9987	2.0000e-06	0.6411	3.9084	0.0751	567.0001	0.9983	27.0	0.9950	3.0	0.9789	chr3	1.2215e+08	Oct4	1.2215e+08	.	NaN	1.3014	NaN	NaN	1.3014	Nanog	1.3014	high	0.9997	NaN	0.6316	NaN	NaN	0.6316	Nanog	0.6316	high	0.7398	Nanog/metacluster_0/p...	485.0	489.0	1.2215e+08	9.0	Nanog	Nanog/m0_p1	480.0	1.2215e+08	5.6011	medium	0.3318	+	Nanog	485.0	-	263220.0	263220.0	4.7505	0.4059	64.0001	0.6206	4.0	0.7477	1.0000e+00	0.7316	2.3451	0.1969	4213.0	0.9990	281.0	0.9987	122.0	0.9973	2.0313	0.0644	2771.0000	0.9993	104.0	0.9987	17.0	0.9856	2.3545	0.0104	795.0001	0.9993	29.0	0.9960	10.0	0.9977	21.0	21.0	++	Nanog<>Nanog	0.3010	2.0899	0.6021
261313	chr3	1.2215e+08	1	Oct4	1.2215e+08	.	NaN	1.3014	NaN	NaN	1.3014	Nanog	1.3014	high	0.9997	NaN	0.6316	NaN	NaN	0.6316	Nanog	0.6316	high	0.7398	Nanog/metacluster_0/p...	485.0	489.0	1.2215e+08	9.0	Nanog	Nanog/m0_p1	480.0	1.2215e+08	5.6011	medium	0.3318	+	Nanog	485.0	-	263220.0	263220.0	4.7505	0.4059	64.0001	0.6206	4.0	0.7477	1.0000e+00	0.7316	2.3451	0.1969	4213.0000	0.9990	281.0	0.9987	122.0	0.9973	2.0313	0.0644	2771.0000	0.9993	104.0	0.9987	1.7000e+01	0.9856	2.3545	0.0104	795.0001	0.9993	29.0	0.9960	10.0	0.9977	chr3	1.2215e+08	Oct4	1.2215e+08	.	NaN	1.1344	NaN	NaN	1.1344	Nanog	1.1344	high	0.9966	NaN	0.5557	NaN	NaN	0.5557	Nanog	0.5557	medium	0.4921	Nanog/metacluster_0/p...	527.0	531.0	1.2215e+08	9.0	Nanog	Nanog/m0_p1	522.0	1.2215e+08	5.6011	medium	0.3318	+	Nanog	527.0	-	263221.0	263221.0	3.9538	0.2452	90.0001	0.7444	4.0	0.7477	2.0000e-06	0.7316	0.3554	0.0081	5494.0	1.0000	248.0	0.9980	211.0	0.9997	0.3638	0.0034	3398.0000	1.0000	98.0	0.9987	39.0	0.9977	0.4669	0.0034	1036.0000	1.0000	33.0	0.9970	16.0	0.9990	42.0	42.0	++	Nanog<>Nanog	2.0899	2.3263	1.0414
261315	chr3	1.2215e+08	1	Oct4	1.2215e+08	.	NaN	1.1344	NaN	NaN	1.1344	Nanog	1.1344	high	0.9966	NaN	0.5557	NaN	NaN	0.5557	Nanog	0.5557	medium	0.4921	Nanog/metacluster_0/p...	527.0	531.0	1.2215e+08	9.0	Nanog	Nanog/m0_p1	522.0	1.2215e+08	5.6011	medium	0.3318	+	Nanog	527.0	-	263221.0	263221.0	3.9538	0.2452	90.0001	0.7444	4.0	0.7477	2.0000e-06	0.7316	0.3554	0.0081	5494.0000	1.0000	248.0	0.9980	211.0	0.9997	0.3638	0.0034	3398.0000	1.0000	98.0	0.9987	3.9000e+01	0.9977	0.4669	0.0034	1036.0000	1.0000	33.0	0.9970	16.0	0.9990	chr3	1.2215e+08	Oct4	1.2215e+08	.	NaN	0.7304	NaN	NaN	0.7304	Nanog	0.7304	high	0.9237	NaN	0.4998	NaN	NaN	0.4998	Nanog	0.4998	low	0.3095	Nanog/metacluster_0/p...	548.0	552.0	1.2215e+08	9.0	Nanog	Nanog/m0_p1	543.0	1.2215e+08	3.9726	low	0.1077	+	Nanog	548.0	-	263222.0	263222.0	3.7226	0.2107	105.0001	0.7890	5.0	0.8316	1.0000e+00	0.7316	0.2655	0.0037	6653.0	1.0000	297.0	0.9993	222.0	1.0000	0.2509	0.0023	3309.0000	1.0000	98.0	0.9987	22.0	0.9909	0.4081	0.0034	1007.0001	1.0000	33.0	0.9970	11.0	0.9977	21.0	21.0	++	Nanog<>Nanog	2.3263	2.3483	1.2304
261316	chr4	1.2554e+08	56	Oct4	1.2553e+08	.	NaN	0.2922	NaN	NaN	0.2922	Nanog	0.2922	low	0.2605	NaN	0.5820	NaN	NaN	0.5820	Nanog	0.5820	medium	0.5842	Nanog/metacluster_0/p...	431.0	435.0	1.2553e+08	9.0	Nanog	Nanog/m0_p1	426.0	1.2553e+08	5.6739	medium	0.3442	+	Nanog	431.0	-	263240.0	263240.0	1.6106	0.0295	385.0001	0.9809	16.0	0.9842	2.0000e+00	0.8601	1.0755	0.0812	891.0001	0.9524	62.0	0.9701	1.0	0.6347	1.4816	0.0386	632.0001	0.9775	23.0	0.9705	8.0000e+00	0.9574	3.8865	0.0748	113.0001	0.9460	4.0	0.8729	6.0	0.9953	chr4	1.2554e+08	Oct4	1.2553e+08	.	NaN	0.5308	NaN	NaN	0.5308	Nanog	0.5308	high	0.7499	NaN	0.6137	NaN	NaN	0.6137	Nanog	0.6137	high	0.6871	Nanog/metacluster_0/p...	470.0	474.0	1.2553e+08	9.0	Nanog	Nanog/m0_p1	465.0	1.2553e+08	7.9618	high	0.7600	+	Nanog	470.0	-	263241.0	263241.0	1.0167	0.0091	528.0001	0.9933	16.0	0.9842	4.0000e+00	0.9440	0.7461	0.0460	1611.0	0.9802	104.0	0.9839	72.0	0.9943	0.7070	0.0084	1242.0000	0.9960	71.0	0.9973	13.0	0.9769	2.3704	0.0104	259.0001	0.9893	18.0	0.9886	1.0	0.8994	39.0	39.0	++	Nanog<>Nanog	0.3010	1.8633	0.8451
261317	chr2	1.7077e+08	76	Oct4	1.7077e+08	.	NaN	0.2731	NaN	NaN	0.2731	Nanog	0.2731	low	0.2096	NaN	0.6504	NaN	NaN	0.6504	Nanog	0.6504	high	0.7995	Nanog/metacluster_0/p...	397.0	402.0	1.7077e+08	9.0	Nanog	Nanog/m0_p1	393.0	1.7077e+08	5.6011	medium	0.3318	-	Nanog	397.0	+	263244.0	263244.0	2.6274	0.0798	150.0001	0.8799	5.0	0.8316	2.0000e+00	0.8601	0.6339	0.0342	630.0001	0.9212	27.0	0.9212	16.0	0.9567	2.0737	0.0668	328.0001	0.9443	11.0	0.9255	5.0000e+00	0.9195	4.0915	0.0839	113.0001	0.9460	5.0	0.9054	2.0	0.9564	chr2	1.7077e+08	Oct4	1.7077e+08	.	NaN	0.1738	NaN	NaN	0.1738	Nanog	0.1738	low	0.0192	NaN	0.4776	NaN	NaN	0.4776	Nanog	0.4776	low	0.2413	Nanog/metacluster_0/p...	446.0	451.0	1.7077e+08	9.0	Nanog	Nanog/m0_p1	442.0	1.7077e+08	3.9044	low	0.1012	-	Nanog	446.0	+	263245.0	263245.0	2.4882	0.0741	200.0001	0.9215	8.0	0.9279	5.0000e+00	0.9611	0.4569	0.0164	1512.0	0.9789	104.0	0.9839	31.0	0.9815	1.3107	0.0268	629.0001	0.9775	22.0	0.9691	7.0	0.9480	2.5963	0.0138	263.0001	0.9899	15.0	0.9826	5.0	0.9936	49.0	49.0	++	Nanog<>Nanog	1.2304	1.5052	0.4771

df.pattern_start_abs_x

261310    1.2215e+08
261313    1.2215e+08
261315    1.2215e+08
             ...    
281036    3.4693e+07
281038    3.4693e+07
281039    1.2162e+08
Name: pattern_start_abs_x, Length: 10182, dtype: float64

features = ['example_chrom', 'pattern_start_abs', 'pattern_end_abs', 'strand']

def suffix(l, suffix):
    return [x +suffix for x in l]

periodicity_dir = Path('/users/avsec/gdrive/projects/chipnexus/data/periodicity/')

(df[['center_diff', 'strand_combination'] + suffix(features, '_x') + suffix(features, '_y')]
 .to_csv(periodicity_dir / 'Nanog<>Nanog.pairs.csv', index=False))