sns.set(style="ticks")
sns.set_context("poster")
from adjustText import adjust_text

f = '../data/HEK293T.duplex.RBP.stat.bed'
df = pd.read_csv(f, header=0, index_col=0, sep='\t')
color_category = []
for i in df.iterrows():
    i = i[1]
    if float(i['pct(cover RRI) 1']) >= 0.05 and float(i['pvalue']) <= 0.01 and float(i['oddsratio']) >= 1.5:
        color_category.append('high')
    elif float(i['pvalue']) <= 0.01 and float(i['oddsratio']) >= 1.5:
        color_category.append('weak')
    else:
        color_category.append('none')
df['enrich level'] = color_category
df.head()

	# cover RRI 1	pct(cover RRI) 1	# all RRI 1	# cover RRI 2	pct(cover RRI) 2	# all RRI 2	obs/exp	log2(obs/exp)	oddsratio	pvalue	enrich level
AGO1	2017	0.039053	51648	1236	0.023931	51648	1.631877	0.706532	1.631877	9.116186e-42	weak
AGO2	1352	0.026177	51648	1088	0.021066	51648	1.242647	0.313417	1.242647	1.375318e-07	none
AGO3	561	0.010862	51648	296	0.005731	51648	1.895270	0.922404	1.895270	1.331809e-19	weak
AGO4	244	0.004724	51648	187	0.003621	51648	1.304813	0.383843	1.304813	6.806662e-03	none
ALKBH5	613	0.011869	51648	439	0.008500	51648	1.396355	0.481666	1.396355	9.296654e-08	none

fig,ax=plt.subplots(figsize=(10,12))

df['size'] = 0.01
sns.scatterplot(x='log2(obs/exp)', y='pct(cover RRI) 1', hue='enrich level', data=df, size='size')

xs = df[df['enrich level']!='none']['log2(obs/exp)']
ys = df[df['enrich level']!='none']['pct(cover RRI) 1']
ss = df[df['enrich level']!='none'].index

texts = []
for x, y, s in zip(xs, ys, ss):
    texts.append(plt.text(x, y, s))
    
# adjust_text(texts, only_move={'text': 'y'})


plt.ylabel('Percentage of interacting regions')

Text(0,0.5,'Percentage of interacting regions')

pd.concat([df, df], axis=1).shape

(48, 24)

pd.DataFrame(df['enrich level'].value_counts())

	enrich level
none	34
weak	8
high	6

1   198
2  2497
3  1274
4   506
5   217
6   349

a = [1,2,3]

a.extend([4,5])
a

[1, 2, 3, 4, 5]

np.random.seed(1234)

import random

random.sample(a, 2)

[4, 5]

from scipy import stats

stats.pearsonr(a, a)

(1.0, 0.0)