def test_ex12():
import pandas as pd
import os
from tcrsampler.sampler import TCRsampler
# fn = 'britanova_chord_blood.csv' # real file
fn = os.path.join('tcrdist','test_files', 'britanova_chord_blood_sample_5000.csv') # test_only file
t = TCRsampler()
t.ref_df = pd.read_csv(fn)
t.build_background()
t.v_freq
t.j_freq
t.vj_freq
t.sample_background(v ='TRBV10-1*01', j ='TRBJ1-1*01',n=3, depth = 1, seed =1, use_frequency= True )
import os
import pandas as pd
from tcrsampler.sampler import TCRsampler
t = TCRsampler()
fn = os.path.join('britanova_chord_blood.csv')
t.ref_df = pd.read_csv(fn)
t.build_background(max_rows=1000)
t.sample(
[['TRBV10-2*01', 'TRBV10-2*01*01', 1], ['TRBV27*01', 'TRBV27*01*01', 4]],
depth=10)
for k, v in t.ref_dict.items():
print(k, v.shape[0])
def test_v_j_freq_estimates():
d = {
'Unnamed: 0': {
0: 0,
1: 1,
2: 2,
3: 3,
4: 4
},
'v_reps': {
0: 'TRBV24-1*01',
1: 'TRBV5-1*01',
2: 'TRBV7-2*01',
3: 'TRBV3-1*01',
4: 'TRBV7-3*01'
},
'j_reps': {
0: 'TRBJ2-1*01',
1: 'TRBJ2-5*01',
2: 'TRBJ2-3*01',
3: 'TRBJ2-5*01',
4: 'TRBJ2-3*01'
},
'cdr3': {
0: 'CATRQDNEQFF',
1: 'CASSLEETQYF',
2: 'CASSLADTQYF',
3: 'CASSQETQYF',
4: 'CASSLAGGTDTQYF'
},
'count': {
0: 252,
1: 166,
2: 113,
3: 98,
4: 89
},
'freq': {
0: 0.0003726818302818776,
1: 0.0002454967612174273,
2: 0.00016711526516608003,
3: 0.00014493182288739684,
4: 0.00013162175752018694
},
'subject': {
0: 'A5-S11.txt',
1: 'A5-S11.txt',
2: 'A5-S11.txt',
3: 'A5-S11.txt',
4: 'A5-S11.txt'
}
}
df = pd.DataFrame(d)
t = TCRsampler()
t.ref_df = df
t.build_background()
assert t.v_occur_freq == {
'TRBV3-1*01': 0.2,
'TRBV5-1*01': 0.2,
'TRBV7-2*01': 0.2,
'TRBV7-3*01': 0.2,
'TRBV24-1*01': 0.2
}
assert t.j_occur_freq == {
'TRBJ2-1*01': 0.2,
'TRBJ2-3*01': 0.4,
'TRBJ2-5*01': 0.4
}
on='subject')
dfd['freq'] = dfd['freq_x'] / dfd['freq_y']
print(dfd[['freq', 'subject']].groupby(['subject']).sum())
# Test that these will work with TCRsampler
from tcrsampler.sampler import TCRsampler
from tcrdist import repertoire_db
ref = repertoire_db.RefGeneSet(db_file='alphabeta_gammadelta_db.tsv')
ref.generate_all_genes()
ref.all_genes
ref.all_genes['human'].keys()
tsd = TCRsampler()
tsd.ref_df = dfd
tsd.build_background()
# find potential missing:
print([x for x in tsd.v_freq.keys()])
print([x for x in tsd.v_freq.keys() if x not in ref.all_genes['human'].keys()])
assert len([
x for x in tsd.v_freq.keys() if x not in ref.all_genes['human'].keys()
]) == 0
print([x for x in tsd.j_freq.keys()])
print([x for x in tsd.j_freq.keys() if x not in ref.all_genes['human'].keys()])
assert len([
x for x in tsd.j_freq.keys() if x not in ref.all_genes['human'].keys()
]) == 0
tsg = TCRsampler()
tsg.ref_df = dfg
