def control_dictionaries_creator():
"""
Create the control dictionary containing the values corresponding to the score of bp and ppt for every control exons
"""
exon_class.set_debug(0)
dir_path = os.path.dirname(os.path.realpath(__file__))
fasterdb = os.path.dirname(os.path.realpath(__file__)).replace(
"src/minimum_free_energy", "data/fasterDB_lite.db")
seddb = os.path.dirname(os.path.realpath(__file__)).replace(
"src/minimum_free_energy", "data/sed.db")
ctrl_dir = dir_path + "/control_dictionaries/"
cnx = sqlite3.connect(fasterdb)
cnx_sed = sqlite3.connect(seddb)
if not os.path.isdir(ctrl_dir):
os.mkdir(ctrl_dir)
exon_type = "CCE"
exon2remove = union_dataset_function.get_exon_regulated_by_sf(
cnx_sed, "down")
ctrl_exon_list = get_control_exon_information(cnx, exon_type, exon2remove)
print("retrieving upstream intron sequence")
list_exon = [
exon_class.ExonClass(cnx, exon[0], exon[1], exon[2])
for exon in ctrl_exon_list
]
print("calculating mfe")
mfe_list_3ss, mfe_list_5ss = function.mfe_calculator(list_exon)
cur_file = open(ctrl_dir + exon_type + "_mfe.py", "w")
cur_file.write("mfe_3ss=" + str(mfe_list_3ss) + "\n")
cur_file.write("mfe_5ss=" + str(mfe_list_5ss) + "\n")
cur_file.close()
def computing_mfe(cnx, df, output):
"""
Add a column mfe 5'ss to the existing dataframe.
:param cnx: (sqlite3 dataframe object) connection to fasterdb
:param df: (pandas dataframe) table of exons
:param output: (str) files were the mfe results will be created
:return: (pandas dataframe) table of exons with mfe data
"""
exon_class_list = []
exon_list = df[["gene_name", "gene_id", "pos"]].values
for exon in exon_list:
exon_class_list.append(
exon_class.ExonClass(cnx, exon[0], exon[1], exon[2]))
mfe_3ss, mfe_5ss = function.mfe_calculator(exon_class_list,
output,
ps=True)
df["mfe_5ss"] = mfe_5ss
return df
def get_mfe_score_list(output, exon_list, name_list):
"""
Calculate or retrieve mfe score
:param output: (string) path where the result will be stored or retrieved
:param exon_list: (list of ExonClass object) List of exons
:param name_list: (string) the name of the exons list
:return: (2 list of floats) list of bp score and list of ppt score
"""
name_store_file = "%s%s_mfe_score.py" % (output, name_list)
print(name_store_file)
if not os.path.isfile(name_store_file):
print("Calculating mfe score using RNAfold")
mfe_3ss, mfe_5ss = function.mfe_calculator(exon_list)
with open(name_store_file, "w") as outfile:
outfile.write("mfe_3ss=" + str(mfe_3ss) + "\n")
outfile.write("mfe_5ss=" + str(mfe_5ss) + "\n")
else:
print("recovering mfe score already stored in %s" % name_store_file)
sys.path.insert(0, output)
stored = __import__("%s_mfe_score" % name_list)
mfe_3ss = stored.mfe_3ss
mfe_5ss = stored.mfe_5ss
return mfe_3ss, mfe_5ss
def get_exon_info(cnx, sedb, fasterdb_file, exon_list, u1_exons, u2_exons):
"""
:param cnx: (sqlite3 connect object) connexion to fasterdb
:param fasterdb_file: (str) an sqlite3 database file
:param sedb: (str) path to sed database
:param exon_list: (list of 2 int) list of exons
:param u1_exons: (list of list of 2 int) list of exons regulated by U1
:param u2_exons: (list of list of 2 int) list of exons regulated by U2
:return: (list of list of value) list of data
"""
dic = {-1: "-", 1: "+"}
cursor = cnx.cursor()
cursor.execute("ATTACH DATABASE ? as sed", (sedb, ))
cursor.execute("ATTACH DATABASE ? as fasterdb", (fasterdb_file, ))
if exon_list is None:
query = """
SELECT t1.id_gene, t1.pos_on_gene, t1.chromosome,
t1.start_on_chromosome, t1.end_on_chromosome,
t2.strand, t3.iupac_exon, t3.upstream_intron_size,
t3.downstream_intron_size
FROM fasterdb.exons as t1, fasterdb.genes as t2, sed.sed as t3
WHERE t3.gene_id = t1.id_gene
AND t3.exon_pos = t1.pos_on_gene
AND t1.id_gene = t2.id
AND t3.exon_type LIKE '%CCE%'
"""
cursor.execute(query)
res = cursor.fetchall()
new_res = []
for exon in res:
exon = list(exon)
exon[3] = int(exon[3]) - 1
cexon = exon_class_bp.ExonClass(cnx, str(exon[0]), exon[0],
exon[1])
exon_data = bp_ppt_calculator([cexon])
mexon = exon_class.ExonClass(cnx, str(exon[0]), exon[0], exon[1])
mfe_5ss, mfe_3ss = mfe_calculator([mexon])
stretch = catch_index_error(stretch_counter([cexon])["T"], 0)
dic_info = {
"GC_content": exon[6].split(";")[4],
"upstream_intron_size": exon[7],
"downstream_intron_size": exon[8],
"UNA_count": catch_index_error(exon_data[8], 0),
"Hbound_count": catch_index_error(exon_data[6], 0),
"good_bp": catch_index_error(exon_data[3], 0),
"MFE_5SS": catch_index_error(mfe_5ss, 0),
"MFE_3SS": catch_index_error(mfe_3ss, 0),
"T_stretch": stretch,
"U1-regulated": is_in(exon[0:2], u1_exons),
"U2-regulated": is_in(exon[0:2], u2_exons),
}
new_res.append(exon[2:5] + ["%s_%s" % (exon[0], exon[1])] + \
["0", dic[exon[5]]] + [str(dic_info)])
return new_res
count = 0
tot = len(exon_list)
result = []
for exon in exon_list:
count += 1
query = """
SELECT t1.chromosome, t1.start_on_chromosome,
t1.end_on_chromosome, t2.strand, t3.iupac_exon,
t3.upstream_intron_size, t3.downstream_intron_size
FROM fasterdb.exons as t1, fasterdb.genes as t2, sed.sed as t3
WHERE t3.gene_id = t1.id_gene
AND t3.exon_pos = t1.pos_on_gene
AND t1.id_gene = t2.id
AND t3.gene_id = %s
AND t3.exon_pos = %s
""" % (exon[0], exon[1])
cursor.execute(query)
res = cursor.fetchall()
if len(res) > 1:
raise IndexError("Error only one row shoud be return for %s" %
exon)
tmp = list(res[0])
tmp[1] = int(tmp[1]) - 1
cexon = exon_class_bp.ExonClass(cnx, str(exon[0]), exon[0], exon[1])
exon_data = bp_ppt_calculator([cexon])
mexon = exon_class.ExonClass(cnx, str(exon[0]), exon[0], exon[1])
mfe_5ss, mfe_3ss = mfe_calculator([mexon])
stretch = catch_index_error(stretch_counter([cexon])["T"], 0)
dic_info = {
"GC_content": tmp[4].split(";")[4],
"upstream_intron_size": tmp[5],
"downstream_intron_size": tmp[6],
"UNA_count": catch_index_error(exon_data[8], 0),
"Hbound_count": catch_index_error(exon_data[6], 0),
"good_bp": catch_index_error(exon_data[3], 0),
"MFE_5SS": catch_index_error(mfe_5ss, 0),
"MFE_3SS": catch_index_error(mfe_3ss, 0),
"T_stretch": stretch,
"U1-regulated": is_in(exon[0:2], u1_exons),
"U2-regulated": is_in(exon[0:2], u2_exons),
}
exon_data = tmp[0:3] + ["%s_%s" % (exon[0], exon[1])] + \
["0", dic[tmp[3]]] + [str(dic_info)]
result.append(exon_data)
sys.stdout.write("Processing %s/%s\t\t\t\r" % (count, tot))
sys.stdout.flush()
return result