combined_replicates.py

from glob import glob
import json
import os
import subprocess
import sys

import pandas as pd

sys.path.append(os.getcwd())
from config import CONFIG
from python.hmmer_manager import HMMerManager
from python.tcr_clusterer import TCRClusterer
from python.tcr_dist import TCRDist
from python.tcr_multi_clusterer import TCRMultiClusterer
from python.tcr_scorer import TCRScorer
from python.utils import get_df_from_file

def write_full_replicate_dataset(output_filename, reptag, chain="B"):
    repfiles = sorted(glob('{}{}_*_{}.tcrs'.format(CONFIG["IEL_DATA_DIR"], reptag, chain)))
    dfs = []
    for repfile in repfiles:
        dfs.append(get_df_from_file(repfile))
    full_df = pd.concat(dfs)
    full_df.iloc[:, [0, 1]].to_csv(output_filename, index=False, header=False)

cd4_reptag = "CD4"
dn_reptag = "DN"
chain = "B"

dn_filename = "all_dn.csv"
cd4_filename = "all_cd4.csv"
dn_file = os.path.join(CONFIG["TMP_OUTPUT"], dn_filename)
cd4_file = os.path.join(CONFIG["TMP_OUTPUT"], cd4_filename)
write_full_replicate_dataset(dn_file, "DN")
write_full_replicate_dataset(cd4_file, "CD4")

multi_clusterer = TCRMultiClusterer(file_1=cd4_file, file_2=dn_file, species="mouse", outdir=os.path.join(CONFIG["HMM_OUTPUT"], "cd4_dn"))

pd.DataFrame(multi_clusterer.result).transpose().to_csv(os.path.join(CONFIG["TMP_OUTPUT"], 'multiple_clusters.csv'))