diff --git a/src/figures_utils/exons_interactions.py b/src/figures_utils/exons_interactions.py
index ea6042e9a5d6fba54f9f5806f03fe5c3bca35539..fa9087460b6c90f11b2d640bf2643f07978ec104 100644
--- a/src/figures_utils/exons_interactions.py
+++ b/src/figures_utils/exons_interactions.py
@@ -11,6 +11,8 @@ from .config_figures import Config
import seaborn as sns
import random
import pandas as pd
+import doctest
+from typing import List, Tuple, Dict
SF = "HNRNPC"
@@ -28,14 +30,17 @@ def get_info_from_database(cnx: sqlite3.Connection, query: str) -> list:
:param query: the SQL query that allows us to get data from the database
:return list_exons: the list of the ID of all fasterdb exons, e.g:
['9995_6', '9995_7', '9996_1', '9996_10']
+
+ >>> get_info_from_database(sqlite3.connect(Config.db_file),
+ ... 'SELECT id_sample FROM cin_projects LIMIT 2')
+ ['GSM1018961_GSM1018962 ', 'GSM1018963_GSM1018964']
"""
cursor = cnx.cursor()
cursor.execute(query)
res = cursor.fetchall()
list_data = []
for data in res:
- data = str(data).lstrip("('").rstrip("',)")
- list_data.append(data)
+ list_data += list(data)
return list_data
@@ -61,7 +66,7 @@ def get_projects_links_to_a_splicing_factor(cnx: sqlite3.Connection,
return id_project
-def get_ase_events(cnx: sqlite3.Connection, id_project: str) -> list:
+def get_ase_events(cnx: sqlite3.Connection, id_project: str) -> List:
"""
Get every exons regulated (down or up) according to a particular project.
@@ -99,7 +104,7 @@ def get_ase_events(cnx: sqlite3.Connection, id_project: str) -> list:
return nres
-def washing_events(exon_list: list) -> list:
+def washing_events(exon_list: List) -> List:
"""
Remove redundant exons or remove exons showing different regulation.
@@ -139,7 +144,7 @@ def washing_events(exon_list: list) -> list:
def get_every_events_4_a_sl(cnx: sqlite3.Connection, sf_name: str,
- regulation: str) -> tuple:
+ regulation: str) -> Tuple:
"""
Get every splicing events for a given splicing factor.
@@ -172,7 +177,52 @@ def get_every_events_4_a_sl(cnx: sqlite3.Connection, sf_name: str,
# ##########GET ALL EXONS IN INTERACTION (for all ChIA-PET datasets)###########
-def get_exons_in_interaction(cnx: sqlite3.Connection, id_project: str) -> dict:
+def get_co_localisation_by_project(cnx: sqlite3.Connection, id_project: str) \
+ -> List[Tuple[str, str]]:
+ """
+ Return The list of exons that are co-localized.
+
+ :param cnx: Connection to chi-pet database.
+ :param id_project: The id of a project
+ :return: The exons co-localized 2 by 2.
+ """
+ cursor = cnx.cursor()
+ query = """SELECT exon1, exon2
+ FROM cin_exon_interaction
+ WHERE id_project = ?"""
+ cursor.execute(query, (id_project,))
+ res = cursor.fetchall()
+ cursor.close()
+ return res
+
+
+def get_colocalized_exons(interaction: List[Tuple[str, str]], id_project: str
+ ) -> Dict[str, List]:
+ """
+ Get a dictionary linking each exon to it's neighbors.
+
+ :param interaction: A list of co-localized exons
+ :param id_project: The id of a project
+ :return: dictionary linking each exon to it's neighbors.
+
+ >>> interaction = [('1_1', '2_1'), ('1_1', '3_1'), ('3_1', '2_1')]
+ >>> res = {'1_1.test': ['2_1', '3_1'], '2_1.test': ['1_1', '3_1'],
+ ... '3_1.test': ['1_1', '2_1']}
+ >>> get_colocalized_exons(interaction, id_project = 'test') == res
+ True
+ """
+ exons_interactions = {}
+ for key, value in interaction:
+ for k, v in [(key, value), (value, key)]:
+ k = k + "." + id_project
+ if k not in exons_interactions:
+ exons_interactions[k] = [v]
+ else:
+ exons_interactions[k].append(v)
+ return exons_interactions
+
+
+def get_exons_in_interaction(cnx: sqlite3.Connection, id_project: str) -> Dict:
"""
Allows:
- to retrieve in the database the pairs of interacting exons, according to
@@ -188,24 +238,39 @@ def get_exons_in_interaction(cnx: sqlite3.Connection, id_project: str) -> dict:
exons with which it interacts, e.g: '9843_10.GSM1234': ['9843_2', '9843_3',
'9843_4'], according to a ChIA-PET dataset.
"""
- cursor = cnx.cursor()
- query = """SELECT exon1, exon2
- FROM cin_exon_interaction
- WHERE id_project = ?"""
- cursor.execute(query, (id_project,))
- res = cursor.fetchall()
+ interaction = get_co_localisation_by_project(cnx, id_project)
+ return get_colocalized_exons(interaction, id_project)
- exons_interactions = {}
- for key, value in res:
- key = key + "." + id_project
- if key not in exons_interactions:
- exons_interactions[key] = [value]
+
+def combine_project_dictionary(projects_dictionary: Dict[str, List]):
+ """
+ Combine the dictionary of interactions of many projects into one.
+
+ :param projects_dictionary: dictionary of interactions of many projects
+ :return: The combined dictionary
+
+ >>> res = {'1_1.test': ['2_1', '3_1'], '2_1.test': ['1_1', '3_1'],
+ ... '3_1.test': ['1_1', '2_1'], '1_1.test2': ['2_1', '3_1', '4_1'],
+ ... '2_1.test2': ['1_1', '6_1'], '3_1.test2': ['1_1', '2_1'],
+ ... '51_1.test3': ['17_1']}
+ >>> expected = {'1_1': ['2_1', '3_1', '2_1', '3_1', '4_1'],
+ ... '2_1': ['1_1', '3_1', '1_1', '6_1'],
+ ... '3_1': ['1_1', '2_1', '1_1', '2_1'],
+ ... '51_1': ['17_1']}
+ >>> combine_project_dictionary(res) == expected
+ True
+ """
+ final_exons_interactions = {}
+ for key, value in projects_dictionary.items():
+ key = key.split(".")[0]
+ if key not in final_exons_interactions:
+ final_exons_interactions[key] = value
else:
- exons_interactions[key].append(value)
- return exons_interactions
+ final_exons_interactions[key] += value
+ return final_exons_interactions
-def get_every_exons_in_interaction() -> dict:
+def get_every_exons_in_interaction() -> Dict[str, List]:
"""
Allows to:
- obtain for each dataset of ChIA-PET, a dictionary, with for each exon
@@ -230,19 +295,12 @@ def get_every_exons_in_interaction() -> dict:
id_chia_pet)
dict_interaction.update(inter)
- final_exons_interactions = {}
- for key, value in dict_interaction.items():
- key = key.split(".")[0]
- if key not in final_exons_interactions:
- final_exons_interactions[key] = value
- else:
- final_exons_interactions[key].extend(value)
- return final_exons_interactions
+ return combine_project_dictionary(dict_interaction)
# ##############PRODUCE DATA FOR THE CONTROLs PART OF THE FIGURE###############
-def find_controls_exons_in_interaction(final_exons_interactions: dict,
- list_exons_ctrl: list):
+def find_controls_exons_in_interaction(final_exons_interactions: Dict,
+ list_exons_ctrl: List):
"""
- For each exon control of the list list_exons_ctrl it allows to obtain the
complete list of exons with which it interacts, for all the datasets that
@@ -259,24 +317,37 @@ def find_controls_exons_in_interaction(final_exons_interactions: dict,
['15919_3', '7546_8', '4946_3', '3158_10'].
:return nb_exons_ctrl_int: the number of exons control in interaction with
at least one other exon control.
+
+ >>> inter_dic = {'1_1': ['2_1', '3_1'],
+ ... '3_1': ['2_1', '1_1'],
+ ... '5_1': ['1_1', '6_1']}
+ >>> list_exons_ctrl = ['1_1', '3_1', '5_1']
+ >>> find_controls_exons_in_interaction(inter_dic, list_exons_ctrl)
+ 3
+ >>> list_exons_ctrl = ['6_1', '1_1', '3_1']
+ >>> find_controls_exons_in_interaction(inter_dic, list_exons_ctrl)
+ 2
+ >>> inter_dic = {'1_1': ['2_1', '3_1'],
+ ... '3_1': ['7_1', '19_1'],
+ ... '5_1': ['17_1', '6_1']}
+ >>> find_controls_exons_in_interaction(inter_dic, list_exons_ctrl)
+ 1
"""
- ctrl_final_inter = {}
- for exons in list_exons_ctrl:
- for key, value in final_exons_interactions.items():
- if exons == key:
- ctrl_final_inter[key] = value
-
- exons_ctrl_interact_exons_ctrl = {}
- for key_a, value_a in ctrl_final_inter.items():
- for elmt_a in value_a:
- for elmt_b in list_exons_ctrl:
- if elmt_a == elmt_b:
- exons_ctrl_interact_exons_ctrl[key_a] = value_a
- nb_exons_ctrl_int = len(exons_ctrl_interact_exons_ctrl.keys())
- return nb_exons_ctrl_int
-
-
-def random_exons_ctrl(sf_reg: dict, number_exons_ctrl: str) -> float:
+ count = 0
+ for exon in list_exons_ctrl:
+ tmp = [e for e in list_exons_ctrl if e != exon]
+ try:
+ for iexon in final_exons_interactions[exon]:
+ if iexon in tmp:
+ count += 1
+ break
+ except KeyError:
+ continue
+ return count
+
+
+def random_exons_ctrl(final_exons_interactions: Dict[str, List],
+ sf_reg: Dict, number_exons_ctrl: str) -> float:
"""
Allows to:
- Generate a list of exons control of size identical to that of the
@@ -286,6 +357,10 @@ def random_exons_ctrl(sf_reg: dict, number_exons_ctrl: str) -> float:
determine the percentage of exons control in interaction with at least one
other exon control. Then we average the 3 percentages obtained.
+ :param final_exons_interactions: a dictionary with for each exon the
+ complete list of exons with which it interacts, for all the datasets that
+ we have, e.g: '4448_11': ['4448_20'], '4857_5': ['4878_1', '494_6'], see
+ get_every_exons_in_interaction().
:param sf_reg: a dictionary with a list of regulated exons depending on a
splicing factor and its regulation, e.g: {'PTBP1_up': ['345_38', '681_2',
'781_4', '1090_16', '1291_12']}, see get_every_events_4_a_sl().
@@ -306,22 +381,22 @@ def random_exons_ctrl(sf_reg: dict, number_exons_ctrl: str) -> float:
for elmt in value:
list_exons.remove(elmt)
- i = 0
list_percentage_ctrl = []
- while i < 3:
+ for _ in range(3): # Put the number of iteration with function parameters
list_exons_ctrl = random.choices(list_exons, k=number_exons_ctrl)
nb_exons_ctrl_int = find_controls_exons_in_interaction(
- get_every_exons_in_interaction(), list_exons_ctrl)
+ final_exons_interactions, list_exons_ctrl)
percentage_ctrl = (nb_exons_ctrl_int / number_exons_ctrl) * 100
rounded_percentage_ctrl = round(percentage_ctrl, 2)
list_percentage_ctrl.append(rounded_percentage_ctrl)
- i += 1
- average_percent_ctrl = sum(list_percentage_ctrl)/len(list_percentage_ctrl)
+
+ average_percent_ctrl = sum(list_percentage_ctrl) / \
+ len(list_percentage_ctrl)
return average_percent_ctrl
# ####KEPT INTERACTIONS FOR EXONS DOWN OR UP REGULATED BY THE SF STUDIED#####
-def kept_interest_exons_sf(final_exons_interactions: dict, sf_reg: dict,
+def kept_interest_exons_sf(final_exons_interactions: Dict, sf_reg: Dict,
number_exons_sf: str) -> float:
"""
- For each exon (down and in a second time up) regulated by the SF of
@@ -349,24 +424,29 @@ def kept_interest_exons_sf(final_exons_interactions: dict, sf_reg: dict,
:return rounded_percentage_sf: the percentage of exons (down and in a
second time up) regulated by the SF of interest in interaction with at
least one other exon regulated by the same SF, in the same way.
+
+ >>> inter_dic = {'1_1': ['2_1', '3_1'],
+ ... '3_1': ['2_1', '1_1'],
+ ... '5_1': ['1_1', '6_1']}
+ >>> list_exons_ctrl = {'SRSF1_down': ['1_1', '3_1', '5_1']}
+ >>> kept_interest_exons_sf(inter_dic, list_exons_ctrl, 'SRSF1_down_3')
+ 100.0
+ >>> inter_dic = {'1_1': ['2_1', '3_1'],
+ ... '3_1': ['2_1', '1_1'],
+ ... '5_1': ['96_7', '6_1']}
+ >>> list_exons_ctrl = {'SRSF1_down': ['1_1', '3_1', '5_1']}
+ >>> kept_interest_exons_sf(inter_dic, list_exons_ctrl, 'SRSF1_down_3')
+ 66.67
+ >>> inter_dic = {'1_1': ['2_1', '3_1'],
+ ... '3_1': ['2_1', '14_18'],
+ ... '5_1': ['96_7', '6_1']}
+ >>> list_exons_ctrl = {'SRSF1_down': ['1_1', '3_1', '5_1']}
+ >>> kept_interest_exons_sf(inter_dic, list_exons_ctrl, 'SRSF1_down_3')
+ 33.33
"""
- sf_final_inter = dict()
- for key_a, value_a in final_exons_interactions.items():
- key_a = key_a.split(".")[0]
- for key_b, value_b in sf_reg.items():
- for elmt_b in value_b:
- if key_a == elmt_b:
- sf_final_inter[key_a] = value_a
-
- exons_sf_interact_exons_sf = {}
- for key_c, value_c in sf_final_inter.items():
- for elmt_c in value_c:
- for key_d, value_d in sf_reg.items():
- for elmt_d in value_d:
- if elmt_c == elmt_d:
- exons_sf_interact_exons_sf[key_c] = value_c
-
- nb_exons_sf_int = len(exons_sf_interact_exons_sf.keys())
+ nb_exons_sf_int = \
+ find_controls_exons_in_interaction(final_exons_interactions,
+ sf_reg[list(sf_reg.keys())[0]])
number_exons_sf = int(number_exons_sf.split("_")[-1])
percentage_sf = (nb_exons_sf_int / number_exons_sf) * 100
rounded_percentage_sf = round(percentage_sf, 2)
@@ -394,7 +474,8 @@ def launch_figures_creation():
sf_reg_do,
number_exons_do)
# CTRL_DOWN
- average_percent_ctrl_do = random_exons_ctrl(sf_reg_do, number_exons_do)
+ average_percent_ctrl_do = random_exons_ctrl(final_exons_interactions,
+ sf_reg_do, number_exons_do)
# SF_UP
sf_reg_up, number_exons_up = get_every_events_4_a_sl(sqlite3.connect(
@@ -403,7 +484,8 @@ def launch_figures_creation():
sf_reg_up,
number_exons_up)
# CTRL_UP
- average_percent_ctrl_up = random_exons_ctrl(sf_reg_up, number_exons_up)
+ average_percent_ctrl_up = random_exons_ctrl(final_exons_interactions,
+ sf_reg_up, number_exons_up)
# CREATE FIGURE
df = pd.DataFrame({"A": [f"{SF}_down", "Control_down", f"{SF}_up",
@@ -418,4 +500,4 @@ def launch_figures_creation():
if __name__ == "__main__":
- launch_figures_creation()
+ doctest.testmod()