Correlation between loop and other modalities

Contents

Correlation between loop and other modalities#

import cooler
import numpy as np
import pandas as pd
from scipy.sparse import triu
from scipy.stats import pearsonr, zscore, norm
from multiprocessing import Pool
from concurrent.futures import ProcessPoolExecutor, as_completed
from ALLCools.mcds import MCDS
from ALLCools.mcds.utilities import calculate_posterior_mc_frac
import matplotlib as mpl
import matplotlib.pyplot as plt
from matplotlib.colors import LogNorm
from matplotlib import cm as cm
import seaborn as sns

mpl.style.use('default')
mpl.rcParams['pdf.fonttype'] = 42
mpl.rcParams['ps.fonttype'] = 42
mpl.rcParams['font.family'] = 'sans-serif'
mpl.rcParams['font.sans-serif'] = 'Helvetica'

leg = ['L23_IT', 'L4_IT', 'L5_IT', 'L6_IT', 'L6_IT_Car3', 'L56_NP', 'L6_CT', 'L6b', 'L5_ET', 'Amy', 
       'Lamp5', 'Lamp5_LHX6', 'Sncg', 'Vip', 'Pvalb', 'Pvalb_ChC', 'Sst', 'CHD7', 
       'MSN_D1', 'MSN_D2', 'Foxp2', 'SubCtx', 
       'ASC', 'ODC', 'OPC', 'MGC', 'PC', 'EC', 'VLMC'
      ]
legname = ['L2/3-IT', 'L4-IT', 'L5-IT', 'L6-IT', 'L6-IT-Car3', 'L5/6-NP', 'L6-CT', 'L6b', 'L5-ET', 'Amy-Exc', 
       'Lamp5', 'Lamp5-Lhx6', 'Sncg', 'Vip', 'Pvalb', 'Pvalb-ChC', 'Sst', 'Chd7', 
       'MSN-D1', 'MSN-D2', 'Foxp2', 'SubCtx-Cplx', 
       'ASC', 'ODC', 'OPC', 'MGC', 'PC', 'EC', 'VLMC'
      ]
leg2name = {xx:yy for xx,yy in zip(leg, legname)}

leg = {'exc': ['L23_IT', 'L4_IT', 'L5_IT', 'L6_IT', 'L6_IT_Car3', 'L56_NP', 'L6_CT', 'L6b', 'Amy'], 
       'inh': ['Lamp5', 'Lamp5_LHX6', 'Sncg', 'Vip', 'Pvalb', 'Pvalb_ChC', 'Sst', 'CHD7'], 
       'msn': ['MSN_D1', 'MSN_D2', 'Foxp2'], 
       'sub': ['SubCtx'], 
       'glia': ['ASC', 'ODC', 'OPC'], 
       'mgc': ['MGC'], 
       'smc': ['PC'], 
       'endo': ['EC'], 
       'fibro': ['VLMC'],
      }
leg['neu'] = leg['exc'] + leg['inh'] + leg['msn'] + leg['sub']
leg['all'] = leg['neu'] + leg['glia'] + leg['mgc'] + leg['smc'] + leg['endo'] + leg['fibro']

group_name = 'neu'

leg = pd.Index(leg[group_name])
legname = leg.map(leg2name)
res = 10000

indir = '/home/jzhou_salk_edu/sky_workdir/hba/loop_majortype/'
outdir = f'/home/jzhou_salk_edu/sky_workdir/hba/loop_majortype/diff/{group_name}/'

chrom_size_path = f'{indir}hg38_with_chrl.main.chrom.sizes'
chrom_sizes = cooler.read_chromsizes(chrom_size_path, all_names=True)

loopall = pd.read_hdf(f'{outdir}merged_loop.hdf', key='data')
loopall

	0	1	2	3	4	5	Qanova	Eanova	Tanova	mCG_corr	mCH_corr	ATAC_corr
0	chr1	900000	910000	chr1	960000	970000	3.750881	6.097476	2.068213	0.031954	0.094070	0.253426
1	chr1	900000	910000	chr1	970000	980000	3.322128	6.001146	2.007495	-0.090257	0.070470	0.265731
2	chr1	910000	920000	chr1	970000	980000	3.293559	5.439024	2.229271	-0.018746	-0.044665	0.376666
3	chr1	910000	920000	chr1	980000	990000	2.704021	5.648575	2.289167	-0.105304	-0.060823	0.306052
4	chr1	910000	920000	chr1	990000	1000000	2.819877	5.675182	1.669268	-0.354732	-0.320624	0.051738
...	...	...	...	...	...	...	...	...	...	...	...	...
2873610	chr22	50570000	50580000	chr22	50670000	50680000	1.646674	11.822375	1.625390	0.009941	-0.277535	0.078601
2873611	chr22	50580000	50590000	chr22	50670000	50680000	2.256175	11.555016	1.335000	-0.049832	-0.312835	0.082439
2873612	chr22	50590000	50600000	chr22	50670000	50680000	3.531459	11.165133	1.195543	0.075519	-0.135215	0.085896
2873613	chr22	50600000	50610000	chr22	50670000	50680000	4.896728	11.926161	2.028210	0.152859	-0.214300	0.119297
2873614	chr22	50610000	50620000	chr22	50670000	50680000	5.475712	13.031472	3.153314	-0.041111	-0.198155	0.250070

2873615 rows × 12 columns

loopq = pd.read_hdf(f'{outdir}loop_Q.hdf', key='data')
loopt = pd.read_hdf(f'{outdir}loop_T.hdf', key='data')