From 3a733e5958353871bb3f7182281bd8940276a11c Mon Sep 17 00:00:00 2001
From: alapendr <audrey.lapendry@ens-lyon.fr>
Date: Thu, 17 Sep 2020 14:15:56 +0200
Subject: [PATCH] features_interactions.py: modifications to add the distance
between genes or exons in pairwise interactions
---
.../interactions/features_interactions.py | 84 ++++++++++++++++++-
1 file changed, 83 insertions(+), 1 deletion(-)
diff --git a/src/db_utils/interactions/features_interactions.py b/src/db_utils/interactions/features_interactions.py
index 6369dbb8..92253d8b 100644
--- a/src/db_utils/interactions/features_interactions.py
+++ b/src/db_utils/interactions/features_interactions.py
@@ -30,6 +30,7 @@ import re
import numpy as np
import os.path
import sys
+import sqlite3
from ..populate_database import populate_df
@@ -251,12 +252,91 @@ def filtering_2(df_filter_2: pd.DataFrame):
if FILE_1 == FILE_2:
df_filter_2["id_project"] = FILE_1
df_filter_2["id2"] = df_filter_2["id"].astype(str) + "_" + \
- df_filter_2["id_project"].astype(str)
+ df_filter_2["id_project"].astype(str)
del df_filter_2["id"]
df_filter_2.rename(columns={"id2": "id"}, inplace=True)
return df_filter_2
+def get_info_from_database(cnx: sqlite3.Connection, query: str) -> \
+ pd.DataFrame:
+ """
+ Get the exons and the genes information from the database, for example:
+ - for the exons: 1_1 18 28681865 28682388
+ - for the genes: 1 18 28645943 28682388
+
+ :param cnx: connexion to the ChIA-PET database
+ :param query: the SQL query that allows us to get data from the database
+ :return df_res: the dataframe with the data obtained
+ """
+ cursor = cnx.cursor()
+ cursor.execute(query)
+ res = list(cursor.fetchall())
+ df_res = pd.DataFrame(res, columns=["ID", "chr", "start", "stop"])
+ return df_res
+
+
+def add_info_distance_between_features(df: pd.DataFrame) -> pd.DataFrame:
+ """
+ Allows to calculate the distance between the two genes or the two exons
+ studied in interaction and to add this information in the result dataframe.
+ If the result is NULL it is because we study two exons or genes located in
+ the same chromosome.
+ If the result is (null), so NaN, it is because we study two exons or genes
+ which have different identifiers, but strictly identical coordinates.
+
+ :param df: Result of the "filtering_2" function
+ :return df: df with distances added or NULL or (null) see before for more
+ details.
+ """
+ info_exon = get_info_from_database(sqlite3.connect(Config.db_file),
+ """SELECT id, chromosome, start, stop
+ FROM cin_exon""")
+ info_gene = get_info_from_database(sqlite3.connect(Config.db_file),
+ """SELECT id, chromosome, start, stop
+ FROM cin_gene""")
+ if OPT_EXON == "on":
+ df = df.merge(info_exon, left_on="exon1", right_on="ID")
+ df = df.rename(columns={'ID': 'ID1', 'chr': 'chr1', 'start': 'start1',
+ 'stop': 'stop1'})
+ df = df.merge(info_exon, left_on="exon2", right_on="ID")
+ df = df.rename(columns={'ID': 'ID2', 'chr': 'chr2', 'start': 'start2',
+ 'stop': 'stop2'})
+ df.loc[df["start1"] < df["start2"], "distance"] = \
+ (df["start2"] - df["stop1"] + 1)
+ df.loc[df["stop1"] < df["stop2"], "distance"] = \
+ (df["start2"] - df["stop1"] + 1)
+ df.loc[df["start1"] > df["start2"], "distance"] = \
+ (df["start1"] - df["stop2"] + 1)
+ df.loc[df["stop1"] > df["stop2"], "distance"] = \
+ (df["start1"] - df["stop2"] + 1)
+ df.loc[df["chr1"] != df["chr2"], "distance"] = "NULL"
+ df.drop(["start1", "ID1", "chr1", "stop1", "ID2", "chr2", "start2",
+ "stop2"], axis='columns', inplace=True)
+ elif OPT_GENE == "on":
+ info_gene["ID"] = info_gene["ID"].astype(str)
+ df = df.merge(info_gene, left_on="gene1", right_on="ID")
+ df = df.rename(columns={'ID': 'ID1', 'chr': 'chr1', 'start': 'start1',
+ 'stop': 'stop1'})
+ df = df.merge(info_gene, left_on="gene2", right_on="ID")
+ df = df.rename(columns={'ID': 'ID2', 'chr': 'chr2', 'start': 'start2',
+ 'stop': 'stop2'})
+ df.loc[df["start1"] < df["start2"], "distance"] = \
+ (df["start2"] - df["stop1"] + 1)
+ df.loc[df["stop1"] < df["stop2"], "distance"] = \
+ (df["start2"] - df["stop1"] + 1)
+ df.loc[df["start1"] > df["start2"], "distance"] = \
+ (df["start1"] - df["stop2"] + 1)
+ df.loc[df["start1"] > df["start2"], "distance"] = \
+ (df["start1"] - df["stop2"] + 1)
+ df.loc[df["stop1"] > df["stop2"], "distance"] = \
+ (df["start1"] - df["stop2"] + 1)
+ df.loc[df["chr1"] != df["chr2"], "distance"] = "NULL"
+ df.drop(["start1", "ID1", "chr1", "stop1", "ID2", "chr2",
+ "start2", "stop2"], axis='columns', inplace=True)
+ return df
+
+
def create_interaction_table(logging_level: str = "DISABLE"):
"""
Create the interaction tables.
@@ -281,6 +361,8 @@ def create_interaction_table(logging_level: str = "DISABLE"):
logging.debug("Sum weight of identical interaction")
df = filtering_2(df)
logging.debug(df.head())
+ df = add_info_distance_between_features(df)
+ logging.debug(df.head())
df.to_csv(MY_OUT_FILE, index=False, sep="\t", header=True)
if OPT_EXON == "on":
logging.debug("Filling cin_exon_interaction")
--
GitLab