def test_mixcr_to_tcrdist_on_clones():
test_clones = os.path.join('tcrdist', 'test_files_compact',
'SRR5130260.1.test.fastq.output.clns.txt')
df = mixcr.mixcr_to_tcrdist2(chain="delta",
organism="human",
clones_fn=test_clones)
assert isinstance(df, pd.DataFrame)
df1 = mixcr.remove_entries_with_invalid_vgene(df,
chain="delta",
organism="human")
assert isinstance(df, pd.DataFrame)
df1['subject'] = 'SRR5130260.1'
tr = TCRrep(cell_df=df1,
organism="human",
chains=['delta'],
db_file='gammadelta_db.tsv')
print(tr.cell_df.shape[0])
tr.infer_cdrs_from_v_gene(chain='delta', imgt_aligned=True)
tr.index_cols = [
'subject', "v_d_gene", 'd_d_gene', 'j_d_gene', 'cdr3_d_nucseq',
'cdr3_d_aa', 'cdr1_d_aa', 'cdr2_d_aa', 'pmhc_d_aa'
]
tr.deduplicate()
assert isinstance(tr.clone_df, pd.DataFrame)
def test_mixcr_integration_with_wrong_chain():
test_clones_fn = os.path.join('tcrdist', 'test_files_compact',
'SRR5130260.1.test.fastq.output.clns.txt')
df = mixcr.mixcr_to_tcrdist2(chain="gamma",
organism="human",
seqs_fn=None,
clones_fn=test_clones_fn)
df2 = mixcr.remove_entries_with_invalid_vgene(df,
chain="gamma",
organism="human")
assert df2.shape[0] == 0
def test_mixcr_integration_with_correct_chain():
test_clones_fn = os.path.join('tcrdist', 'test_files_compact',
'SRR5130260.1.test.fastq.output.clns.txt')
df = mixcr.mixcr_to_tcrdist2(chain="delta",
organism="human",
seqs_fn=None,
clones_fn=test_clones_fn)
assert isinstance(df, pd.DataFrame)
df1 = mixcr.remove_entries_with_invalid_vgene(df,
chain="delta",
organism="human")
assert isinstance(df, pd.DataFrame)
assert df1.shape[0] == 89
def test_convert_minervina_to_mixcr_run_tcrdist(f, my_chain):
fn = os.path.join('tcrdist', 'test_files', f)
df = pd.read_csv(fn, sep="\t")
df['bestVGene'] = df['bestVGene'].apply(lambda s: s + "*00")
df['bestJGene'] = df['bestJGene'].apply(lambda s: s + "*00")
map_minervina_to_mixcr = \
{'Rank':'cloneId',
'Read.count':'cloneCount',
'Read.proportion':'cloneFraction',
'bestVGene': 'allVHitsWithScore',
'bestDGene': 'allDHitsWithScore',
'bestJGene':'allJHitsWithScore',
'CDR3.nucleotide.sequence':'nSeqCDR3',
'CDR3.amino.acid.sequence':'aaSeqCDR3',
'refPoints':'refPoints'}
df = df.rename(columns=map_minervina_to_mixcr)
# CREATE A FAUX MIXCR OUTPUT
df.to_csv('dfmix.clns.txt', index=False, sep="\t")
# USE TCRDIST2 TOOL FOR PORTING MIXCR OUTPUTS
dfmix = mixcr.mixcr_to_tcrdist2(chain=my_chain,
organism="human",
clones_fn='dfmix.clns.txt')
if my_chain == "alpha":
assert set(dfmix.columns.to_list()) == set([
'clone_id', 'count', 'v_a_gene', 'd_a_gene', 'j_a_gene',
'cdr3_a_nucseq', 'cdr3_a_aa'
])
assert df.shape[0] == dfmix.shape[0]
dfmix = mixcr.remove_entries_with_invalid_vgene(dfmix,
chain=my_chain,
organism="human")
dfmix = mixcr.remove_entries_with_invalid_cdr3(dfmix, chain=my_chain)
if my_chain == "alpha":
tr = TCRrep(cell_df=dfmix, organism="human", chains=['alpha'])
elif my_chain == "beta":
tr = TCRrep(cell_df=dfmix, organism="human", chains=['beta'])
tr.infer_cdrs_from_v_gene(chain=my_chain, imgt_aligned=True)
tr.cell_df['subject'] = 'X'
tr.cell_df['epitope'] = 'X'
if my_chain == "alpha":
tr.index_cols = [
'clone_id', 'subject', 'epitope', 'v_a_gene', 'j_a_gene',
'cdr3_a_aa', 'cdr1_a_aa', 'cdr2_a_aa', 'pmhc_a_aa', 'cdr3_a_nucseq'
]
elif my_chain == "beta":
tr.index_cols = [
'clone_id', 'subject', 'epitope', 'v_b_gene', 'j_b_gene',
'cdr3_b_aa', 'cdr1_b_aa', 'cdr2_b_aa', 'pmhc_b_aa', 'cdr3_b_nucseq'
]
tr.deduplicate()
if my_chain == "alpha":
tr._tcrdist_legacy_method_alpha()
assert isinstance(tr.cdr3_a_aa_pw, np.ndarray)
assert isinstance(tr.paired_tcrdist, np.ndarray)
elif my_chain == "beta":
tr._tcrdist_legacy_method_beta()
assert isinstance(tr.cdr3_b_aa_pw, np.ndarray)
assert isinstance(tr.paired_tcrdist, np.ndarray)
def test_combine_betas_and_alphas():
import pytest
import os
import numpy as np
import pandas as pd
from tcrdist.repertoire import TCRrep
from tcrdist import mixcr
import multiprocessing
testfiles = [
('contracting_clones_M_alpha.tsv', 'alpha', 'M', 'contracting'),
('contracting_clones_M_beta.tsv', 'beta', 'M', 'contracting'),
('contracting_clones_W_alpha.tsv', 'alpha', 'W', 'contracting'),
('contracting_clones_W_beta.tsv', 'beta', 'W', 'contracting'),
('expanding_clones_M_alpha.tsv', 'alpha', 'M', 'expanding'),
('expanding_clones_M_beta.tsv', 'beta', 'M', 'expanding'),
('expanding_clones_W_alpha.tsv', 'alpha', 'W', 'expanding'),
('expanding_clones_W_beta.tsv', 'beta', 'W', 'expanding')
]
betas = []
alphas = []
for f, my_chain, sub, group in testfiles:
fn = os.path.join('tcrdist', 'test_files', f)
df = pd.read_csv(fn, sep="\t")
df['bestVGene'] = df['bestVGene'].apply(lambda s: s + "*00")
df['bestJGene'] = df['bestJGene'].apply(lambda s: s + "*00")
print(df.columns)
map_minervina_to_mixcr = \
{'Rank':'cloneId',
'Read.count':'cloneCount',
'Read.proportion':'cloneFraction',
'bestVGene': 'allVHitsWithScore',
'bestDGene': 'allDHitsWithScore',
'bestJGene':'allJHitsWithScore',
'CDR3.nucleotide.sequence':'nSeqCDR3',
'CDR3.amino.acid.sequence':'aaSeqCDR3',
'refPoints':'refPoints'}
df = df.rename(columns=map_minervina_to_mixcr)
# CREATE A FAUX MIXCR OUTPUT
print(df.columns)
df.to_csv('dfmix.clns.txt', index=False, sep="\t")
dfmix = mixcr.mixcr_to_tcrdist2(chain=my_chain,
organism="human",
clones_fn='dfmix.clns.txt')
dfmix['CD'] = df['CD'].copy()
dfmix['proportion'] = df['cloneFraction'].copy()
dfmix = mixcr.remove_entries_with_invalid_vgene(dfmix,
chain=my_chain,
organism="human")
dfmix = mixcr.remove_entries_with_invalid_cdr3(dfmix, chain=my_chain)
dfmix['source'] = f
dfmix['subject'] = sub
dfmix['epitope'] = 'X'
dfmix['trajectory'] = group
if my_chain == "alpha":
alphas.append(dfmix)
elif my_chain == "beta":
betas.append(dfmix)
betas_joined = pd.concat(betas)
alpha_joined = pd.concat(alphas)
testsets = [(alpha_joined, 'alpha'), (betas_joined, 'beta')]
tcr_rep_results = dict()
for dfmix, my_chain in testsets:
if my_chain == "alpha":
tr = TCRrep(cell_df=dfmix, organism="human", chains=['alpha'])
elif my_chain == "beta":
tr = TCRrep(cell_df=dfmix, organism="human", chains=['beta'])
tr.infer_cdrs_from_v_gene(chain=my_chain, imgt_aligned=True)
if my_chain == "alpha":
tr.index_cols = [
'clone_id', 'subject', 'epitope', 'trajectory', 'v_a_gene',
'j_a_gene', 'cdr3_a_aa', 'cdr1_a_aa', 'cdr2_a_aa', 'pmhc_a_aa',
'cdr3_a_nucseq', 'CD', 'source'
]
elif my_chain == "beta":
tr.index_cols = [
'clone_id', 'subject', 'epitope', 'trajectory', 'v_b_gene',
'j_b_gene', 'cdr3_b_aa', 'cdr1_b_aa', 'cdr2_b_aa', 'pmhc_b_aa',
'cdr3_b_nucseq', 'CD', 'source'
]
tr.deduplicate()
my_processes = 10
if my_processes > multiprocessing.cpu_count():
my_processes = multiprocessing.cpu_count()
if my_chain == "alpha":
tr._tcrdist_legacy_method_alpha(processes=my_processes)
assert isinstance(tr.cdr3_a_aa_pw, np.ndarray)
assert isinstance(tr.paired_tcrdist, np.ndarray)
elif my_chain == "beta":
tr._tcrdist_legacy_method_beta(processes=my_processes)
assert isinstance(tr.cdr3_b_aa_pw, np.ndarray)
assert isinstance(tr.paired_tcrdist, np.ndarray)
tcr_rep_results[my_chain] = tr
return tcr_rep_results