if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/iTreg_enhancers/2014_02_14/Motifs'
dirpath = yzer.get_path(dirpath)
for ab in ('me2', 'ac'):
for condition in ('treg', 'itreg', 'activated'):
cond_dirpath = yzer.get_filename(dirpath,
'{}_{}'.format(condition, ab))
if True:
filename = yzer.get_filename(
cond_dirpath, '{}_{}_enhancers.txt'.format(condition, ab))
data = yzer.import_file(filename)
data = data.fillna(0)
yzer.run_homer(data,
'all',
cond_dirpath,
cpus=6,
center=True,
reverse=False,
preceding=False,
size=200,
length=[8, 10, 12, 15],
mock=True)
for condition in ('itreg', 'treg'):
cond_dirpath = yzer.get_filename(dirpath,
'''
Created on Jul 5, 2012
@author: karmel
'''
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
from glasslab.dataanalysis.misc.gr_project_2012.elongation import get_rep_string
import sys
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland_2012/GR_Project/motifs'
dirpath = yzer.get_path(dirpath)
pausing_data = yzer.import_file(
yzer.get_filename(dirpath, 'feature_vectors.txt'))
data = yzer.import_file(
yzer.get_filename(dirpath, 'transcript_vectors.txt'))
try:
min_ratio = float(sys.argv[1])
except IndexError:
min_ratio = 1.5
if False:
yzer.prep_files_for_homer(data,
'all_transcripts_promoter',
dirpath,
center=False,
reverse=False,
preceding=True,
Created on Nov 7, 2012
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.misc.cd4tcell_finland_2012.resources import replicate_sets
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells_Finland_2012/Analysis_2013_02'
dirpath = yzer.get_path(dirpath)
go_path = yzer.get_and_create_path(dirpath, 'with_me3', 'go_analysis',
'0_8_min_lfc')
data = yzer.import_file(
yzer.get_filename(dirpath, 'transcript_vectors.txt'))
data = data.fillna(0)
data = data[data['naive_me3_tag_count'] + data['act_me3_tag_count'] > 0]
if False:
curr_path = yzer.get_and_create_path(dirpath, 'with_me3',
'motif_analysis')
yzer.run_homer(data,
'all_refseq_preceding',
curr_path,
center=False,
reverse=False,
preceding=True,
size=200,
cpus=6)
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
from collections import OrderedDict
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/TReg_enhancers/2013_04_01'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath, 'motifs')
min_score = 10
stats = OrderedDict()
all_data = yzer.import_file(
yzer.get_filename(dirpath, 'foxp3_with_treg_enhancers.txt')).fillna(0)
stats['all'] = len(all_data)
# Filter out promoters
data = all_data[all_data['tss_id'] == 0]
stats['enhancers'] = len(data)
stats['me2_treg'] = sum(data['me2_id'] > 0)
stats['ac_treg'] = sum(data['ac_id'] > 0)
stats['me2_treg_naive'] = sum((data['me2_id'] > 0)
& (data['naive_id'] > 0))
stats['me2_treg_th1'] = sum((data['me2_id'] > 0) & (data['th1_id'] > 0))
stats['me2_treg_naive_th1'] = sum((data['me2_id'] > 0)
& (data['naive_id'] > 0)
& (data['th1_id'] > 0))
stats['me2_treg_only'] = sum((data['me2_id'] > 0)
& (data['naive_id'] == 0)
@author: karmel
The goal here is to see if we can tease out which features are
predictive of characteristics of interest, such as pausing ratio
or transrepression. Worth a stab.
'''
from glasslab.dataanalysis.misc.gr_project_2012.elongation import get_rep_string
import sys
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland_2012/GR_Project/motifs'
dirpath = yzer.get_path(dirpath)
grouped = yzer.import_file(
yzer.get_filename(dirpath, 'feature_vectors.txt'))
paused = []
if True:
# Can we predict pausing ratio?
# Minimal ratio in KLA+Dex vs. KLA pausing
try:
min_ratio = float(sys.argv[1])
except IndexError:
min_ratio = 2
try:
secondary_min_ratio = float(sys.argv[2])
except IndexError:
secondary_min_ratio = 1.2
'''
Created on Sep 7, 2012
@author: karmel
'''
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland_2012/GR_Project/'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath, 'motifs', 'from_peaks')
all_data = yzer.import_file(
yzer.get_filename(motif_dirpath, 'p65_kla_vectors.txt'))
size = 100
if True:
for ratio in (3, 2, 1.5):
enhancers = yzer.import_file(
yzer.get_filename(
dirpath, 'boxplots_non_refseq_by_p65',
'enhancer_like_lose_p65_{0}x_change_dsg_only.txt'.format(
ratio)))
enhancers['glass_transcript_id'] = enhancers['id']
# Limit to peaks and touching transcripts, then pull out peaks that intersect our enhancer set
data = all_data[all_data['touches'] == 't']
data = data.merge(enhancers,
how='right',
on='glass_transcript_id',
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/Foxo1'
dirpath = yzer.get_path(dirpath)
motifs_dirpath = yzer.get_and_create_path(dirpath, 'motifs')
peak_pretty = 'Foxo1'
peak = peak_pretty.lower()
foxo1 = yzer.import_file(
yzer.get_filename(dirpath,
'{0}_with_foxp3.txt'.format(peak))).fillna(0)
datasets = [
('foxo1_all', foxo1),
('foxo1_tss', foxo1[foxo1['tss_id'] > 0]),
('foxo1_enhancers', foxo1[foxo1['tss_id'] == 0]),
]
for name, subset in datasets:
subset['id'] = subset['{0}_id'.format(peak)]
subset['start'] = subset['{0}_start'.format(peak)]
subset['end'] = subset['{0}_end'.format(peak)]
yzer.run_homer(subset,
name,
motifs_dirpath,
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/Rudensky_enhancers'
dirpath = yzer.get_path(dirpath)
motifs_dirpath = yzer.get_and_create_path(dirpath, 'motifs')
peak_pretty = 'Foxp3'
peak = peak_pretty.lower()
foxp3 = yzer.import_file(
yzer.get_filename(dirpath,
'{0}_1_with_naive_me2.txt'.format(peak))).fillna(0)
naive = yzer.import_file(
yzer.get_filename(dirpath,
'naive_me2_with_{0}.txt'.format(peak))).fillna(0)
# Filter out promoters
foxp3 = foxp3[foxp3['tss_id'] == 0]
naive = naive[naive['tss_id'] == 0]
# Get venn-diagram sets for foxp3/me2
only_foxp3 = foxp3[foxp3['naive_id'] == 0]
only_naive = naive[naive['foxp3_1_id'] == 0]
shared = foxp3[foxp3['naive_id'] > 0]
print len(only_foxp3), len(only_naive), len(shared)
'''
Created on Sep 7, 2012
@author: karmel
'''
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland_2012/GR_Project/motifs'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath,'from_peaks')
transcripts = yzer.import_file(yzer.get_filename(dirpath, 'transcript_vectors.txt'))
transcripts['glass_transcript_id'] = transcripts['id']
for peak_type in ('gr_dex', 'gr_kla_dex', 'p65_kla_dex','p65_kla'):
size = 100
if True:
all_data = yzer.import_file(yzer.get_filename(motif_dirpath,
'{0}_vectors.txt'.format(peak_type)))
all_data = all_data.merge(transcripts, how='left', on='glass_transcript_id',suffixes=['','trans'])
all_data = all_data.fillna(0)
for super_name, data in (#('all', all_data,),
#('refseq', all_data[(all_data['score'] > 10) & (all_data['has_refseq'] == 1)
# & (all_data['touches'] == 't') | (all_data['relationship'] == 'is downstream of')],),
#('distal', all_data[(all_data['distal'] == 't')],),
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/H3K4me2/Analysis/venn_diagrams'
dirpath = yzer.get_path(dirpath)
motifs_dirpath = yzer.get_and_create_path(dirpath, 'motifs')
peak_pretty = 'H3K4me2'
peak = peak_pretty.lower()
th1 = yzer.import_file(
yzer.get_filename(
dirpath, 'th1_with_th2_me1_and_{0}.txt'.format(peak))).fillna(0)
th2 = yzer.import_file(
yzer.get_filename(
dirpath, 'th2_with_th1_me1_and_{0}.txt'.format(peak))).fillna(0)
naive = yzer.import_file(
yzer.get_filename(
dirpath, '{0}_with_th1_and_th2_me1.txt'.format(peak))).fillna(0)
# Filter out promoters
th1 = th1[th1['tss_id'] == 0]
th2 = th2[th2['tss_id'] == 0]
naive = naive[naive['tss_id'] == 0]
# Get venn-diagram sets for th1/th2
only_th1 = th1[th1['th2_id'] == 0]
if __name__ == '__main__':
yzer = MotifAnalyzer()
grapher = SeqGrapher()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/TReg_enhancers/2013_04_01'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath, 'motifs')
graph_dirpath = yzer.get_filename(dirpath, 'piecharts')
min_score = 10
for antibody in ('me2', ):
data = {}
celltypes = ['treg', 'naive', 'th1', 'th2']
for celltype in ('treg', 'th1'):
data[celltype] = yzer.import_file(
yzer.get_filename(
dirpath, '{0}_{1}_versus_others_with_foxp3.txt'.format(
celltype, antibody))).fillna(0)
# Filter out promoters
data[celltype] = data[celltype][data[celltype]['tss_id'] == 0]
# Get venn-diagram "only" sets
others = celltypes[:]
others.remove(celltype)
data[celltype + '_only'] = data[celltype][data[celltype]\
[[o + '_id' for o in others]].sum(axis=1) == 0]
# Pairwise
for other in others:
other_pair = others[:]
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland_2012/GR_Project/motifs/from_peaks/vs_non_dsg'
dirpath = yzer.get_path(dirpath)
for peak_type in ( #'gr_dex_fa',
'gr_kla_dex_fa',
#'gr_dex_dsg', 'gr_kla_dex_dsg',
):
size = 200
if True:
all_data = yzer.import_file(
yzer.get_filename(dirpath,
'{0}_vectors.txt'.format(peak_type)))
all_data = all_data.fillna(0)
for super_name, data in ((
'all',
all_data,
), ):
for name, dataset in ((
'all',
data,
), ):
# We have multiple copies of peaks if they align to different transcripts
curr_path = yzer.get_and_create_path(
dirpath, peak_type, super_name, name)
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/H3K4me2/Analysis/venn_diagrams'
dirpath = yzer.get_path(dirpath)
motifs_dirpath = yzer.get_and_create_path(dirpath, 'motifs')
peak_pretty = 'H3K4me2'
peak = peak_pretty.lower()
dp = yzer.import_file(
yzer.get_filename(dirpath, 'dp_with_{0}.txt'.format(peak))).fillna(0)
naive = yzer.import_file(
yzer.get_filename(dirpath, '{0}_with_dp.txt'.format(peak))).fillna(0)
# Filter out promoters
dp = dp[dp['tss_id'] == 0]
naive = naive[naive['tss_id'] == 0]
# Get venn-diagram sets for th1/th2
dp_only = dp[dp['naive_id'] == 0]
naive_only = naive[naive['dp_id'] == 0]
shared = naive[naive['dp_id'] > 0]
datasets = [dp_only, naive_only, shared]
main_peak = ['dp', 'naive', 'naive']
names = [
Created on Apr 18, 2013
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
from glasslab.dataanalysis.graphing.seq_grapher import SeqGrapher
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/TReg_enhancers/2013_04_01'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath, 'motifs')
foxp3 = yzer.import_file(yzer.get_filename(
dirpath, 'foxp3_with_stat1.txt')).fillna(0)
stat1 = yzer.import_file(yzer.get_filename(
dirpath, 'stat1_with_foxp3.txt')).fillna(0)
print len(foxp3), len(stat1)
print sum(foxp3['stat1_id'] > 0)
print sum(foxp3['stat1_id'] > 0) / len(foxp3)
print sum(stat1['foxp3_id'] > 0)
print sum(stat1['foxp3_id'] > 0) / len(stat1)
foxp3_enh = foxp3[(foxp3['tss_me2_id'] == 0) & (foxp3['tss_id'] == 0)]
foxp3_tss = foxp3[(foxp3['tss_me2_id'] > 0) | (foxp3['tss_id'] > 0)]
print len(foxp3_enh)
print sum(foxp3_enh['stat1_id'] > 0) / len(foxp3_enh)
print len(foxp3_tss)
print sum(foxp3_tss['stat1_id'] > 0) / len(foxp3_tss)
Created on Mar 28, 2013
@author: karmel
'''
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/TReg_enhancers/2013_03_19'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath, 'motifs')
for antibody in ('me2', 'ac'):
treg = yzer.import_file(
yzer.get_filename(
dirpath, 'treg_with_naive_{0}.txt'.format(antibody))).fillna(0)
naive = yzer.import_file(
yzer.get_filename(
dirpath, 'naive_with_treg_{0}.txt'.format(antibody))).fillna(0)
# Filter out promoters
treg = treg[treg['tss_id'] == 0]
naive = naive[naive['tss_id'] == 0]
# Get venn-diagram sets for foxp3/me2
only_treg = treg[treg['naive_id'] == 0]
only_naive = naive[naive['treg_id'] == 0]
shared = treg[treg['naive_id'] > 0]
print len(only_treg), len(only_naive), len(shared)
'''
Created on Feb 8, 2013
@author: karmel
'''
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
base_dirpath = yzer.get_path(
'karmel/GlassLab/Notes_and_Reports/NOD_BALBc/ThioMacs/Analysis_2013_02/'
)
dirpath = yzer.get_and_create_path(base_dirpath, 'motifs/')
filename = yzer.get_filename(base_dirpath, 'transcript_vectors.txt')
data = yzer.import_file(filename)
data = data.fillna(0)
# Promoters
if False:
refseq = data[data['has_refseq'] == 1]
refseq = refseq[refseq['transcript_score'] >= 4]
if True:
yzer.run_homer(refseq,
'refseq_promoter',
dirpath,
cpus=6,
center=False,
reverse=False,
preceding=True,
size=400,
Created on Jul 4, 2012
@author: karmel
'''
from glasslab.dataanalysis.misc.gr_project_2012.elongation import get_rep_string
import sys
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
from glasslab.dataanalysis.graphing.seq_grapher import SeqGrapher
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland 2012/KLA skew/'
dirpath = yzer.get_path(dirpath)
grouped = yzer.import_file(
yzer.get_filename(dirpath, 'feature_vectors.txt'))
repressed = []
if True:
# Can we predict pausing ratio?
# Minimal ratio in KLA+Dex vs. KLA pausing
try:
min_ratio = float(sys.argv[1])
except IndexError:
min_ratio = -1
try:
thresh = int(sys.argv[3])
except IndexError:
thresh = 4
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.misc.gr_project_2012.elongation import get_rep_string
import sys
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
from glasslab.dataanalysis.graphing.seq_grapher import SeqGrapher
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland 2012/KLA skew/'
dirpath = yzer.get_path(dirpath)
grouped = yzer.import_file(
yzer.get_filename(dirpath, 'feature_vectors.txt'))
repressed = []
if True:
# Can we predict pausing ratio?
# Minimal ratio in KLA+Dex vs. KLA pausing
try:
min_ratio = float(sys.argv[1])
except IndexError:
min_ratio = -1
try:
thresh = int(sys.argv[3])
except IndexError:
thresh = 4
'''
Created on Apr 12, 2013
@author: karmel
'''
from __future__ import division
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/GlassLab/Notes_and_Reports/CD4TCells/TReg_enhancers/2013_04_01'
dirpath = yzer.get_path(dirpath)
data = yzer.import_file(yzer.get_filename(
dirpath, 'th1_with_stat1_ko.txt')).fillna(0)
print len(data)
data = data[data['tss_id'] == 0]
data['ko_ratio'] = data['ko_id'] / data['th1_id']
data['treg_ratio'] = data['treg_id'] / data['th1_id']
enh = len(data)
print enh
print sum(data['ko_ratio'] < .5), sum(data['ko_ratio'] < .5) / enh
print sum(data['treg_ratio'] < .5), sum(data['treg_ratio'] < .5) / enh
print sum((data['treg_ratio'] < .5) & (data['ko_ratio'] < .5))
print sum((data['treg_ratio'] < .5) & (data['ko_ratio'] < .5)) / enh
'''
Created on Sep 7, 2012
@author: karmel
'''
from glasslab.dataanalysis.motifs.motif_analyzer import MotifAnalyzer
from glasslab.dataanalysis.misc.gr_project_2012.v1.boxplots_redistribution_pairs import get_high_quality_pairs
if __name__ == '__main__':
yzer = MotifAnalyzer()
dirpath = 'karmel/Desktop/Projects/Classes/Rotations/Finland_2012/GR_Project/'
dirpath = yzer.get_path(dirpath)
motif_dirpath = yzer.get_filename(dirpath, 'motifs', 'from_peaks')
transcripts = yzer.import_file(
yzer.get_filename(dirpath, 'motifs', 'transcript_vectors.txt'))
transcripts['glass_transcript_id'] = transcripts['id']
size = 200
if True:
all_data = yzer.import_file(
yzer.get_filename(
dirpath, 'redistribution',
'p65_peaks_bigger_in_kla_dex_with_nearby_bigger_kla_peaks.txt')
)
all_data = get_high_quality_pairs(all_data, transcripts)
for super_name, data in (
(
'all',
