### Made by Hugo Gillet ###
import pandas as pd
import json
import re
import representative as repr
import os
def dict_reprTrans_to_df(dict_reprTrans: dict[str, str]) -> pd.DataFrame:
"""Convert a dictionary of genes and their representative transcript into a dataframe
Args:
dict_reprTrans (dict) : {'Gene':['transcriptA', 'transcriptB'], ...}
Returns:
Pandas dataframe having Gene and transcript as columns
Raises:
Only dict are allowed
Key should be strings
Value should be strings
"""
pass
if not type(dict_reprTrans) is dict:
raise TypeError("Only dict are allowed")
if type(list(dict_reprTrans.keys())[0]) is not str:
raise TypeError("Key should be strings")
if type(list(dict_reprTrans.values())[0]) is not str:
raise TypeError("Values should be strings")
df_reprTrans = pd.DataFrame.from_dict(
dict_reprTrans, orient="index", columns=["reprTranscript"]
)
df_reprTrans = df_reprTrans.reset_index(level=0)
df_reprTrans.columns = ["Gene", "reprTrans"]
df_reprTrans["reprTrans"] = df_reprTrans["reprTrans"].str.replace(
r"\.[1-9]", "", regex=True
)
return df_reprTrans
def txt_to_dict(dict_txt: str) -> dict:
"""Convert a txt file into a dictionary
Args:
dict_txt (str) : pathe to a txt file of a dict
structured as {'Gene':['transcriptA', 'transcriptB'], ...}
Returns:
dict (dict) : dictionary stuctured as {'Gene':['transcriptA', 'transcriptB'], ...}
Raises:
None
"""
pass
input: str = open(dict_txt, "r").read()
input: str = input.replace("'", '"')
dict = json.loads(input)
return dict
def transcripts_by_gene_inDf(df_gtfSelection: pd.DataFrame) -> pd.DataFrame:
"""Convert multiindex dataframe from function into a simple dataframe
Args:
df_gtfSelection : Pandas multiindex dataframe having Gene,
transcript as indexs and support level as columns.
Come from the function import_gtfSelection_to_df()
from representative.py script.
Returns:
df_gene (str): Pandas dataframe having Gene and
transcript as columns
Raises:
None
"""
pass
df_gene = df_gtfSelection.set_index(["Gene"])
df_gene = df_gene.drop(columns=["Support_level"])
df_gene["Transcript"] = df_gene["Transcript"].str.replace(
r"\.[0-9]", "", regex=True
)
df_gene = df_gene.reset_index(level=0)
return df_gene
def tsv_or_csv_to_df(input_txt: str) -> pd.DataFrame:
"""Convert tsv or csv file into a pandas dataframe
Args:
input_txt (str): csv or tsv file containing transcript expression level
Returns:
df_gene (str): Pandas dataframe having transcript and expression level
as columns
Raises:
None
"""
pass
df_input = pd.read_csv(
input_txt,
sep=r"[\t,]",
lineterminator="\n",
names=["Transcript", "Expression_level"],
engine="python",
)
return df_input
def exprLevel_byGene(
df_exprTrasncript: pd.DataFrame, df_output_gtf_selection: pd.DataFrame
) -> pd.DataFrame:
"""find the gene of each transcipt given by the expression level csv/tsv file,
and summ expression level of all transcipts from the same gene.
Args:
df_exprTranscript : pandas Dataframe containing transcript and their expression level,
generated by "tsv_or_csv_to_df" function
df_output_gtf_selection : pandas Dataframe containing genes and transcripts,
generated by "transcripts_by_gene_inDf" function
Returns:
Pandas dataframe having gene and sum of its transcript expression level
Raises:
None
"""
pass
df_merged = pd.merge(
df_output_gtf_selection, df_exprTrasncript, how="inner", on="Transcript"
)
df_sum = df_merged.groupby("Gene").sum(
"Expression_level"
) # sum transcripts comming from the same gene
return df_sum
def match_byGene(
df_reprTranscript: pd.DataFrame, df_expressionLevel_byGene: pd.DataFrame
) -> pd.DataFrame:
"""Find matching genes bewteen the 2 args
Args:
df_reprTranscript : pandas Dataframe containing genes
and their representative transcript, generated by
"dict_reprTrans_to_df()"
df_expressionLevel_byGene : pandas Dataframe containing
genes and their expression level generated by
"transcript_by_gene_inDf()"
Returns:
Pandas dataframe having representative trasncripts
and their expression level
Raises:
None
"""
pass
df_merged = pd.merge(
df_reprTranscript, df_expressionLevel_byGene, how="outer", on="Gene"
)
df_clean = df_merged.dropna(axis=0)
df_clean = df_clean.loc[:, ["reprTrans", "Expression_level"]]
return df_clean
def output_tsv(dataframe: pd.DataFrame) -> pd.DataFrame:
"""Convert pandas dataframe into a tsv file
Args:
dataframe (str): Pandas dataframe containing
representative transcripts and their expression level
Returns:
Tsv file containing representative transcripts
and their expression level in the same directory
Raises:
None
"""
pass
csv_file = dataframe.to_csv(
os.getcwd() + "\ReprTrans_ExpressionLevel.tsv",
sep="\t",
index=False,
header=False,
)
return csv_file
### functions to run this part of the programm
def match_reprTranscript_expressionLevel(
exprTrans: str, dict_reprTrans: dict, intermediate_file: str
):
"""Combine functions to replace transcripts from an expression level csv/tsv file
with representative transcripts
Args:
exprTrans (str): csv or tsv file containing transcripts
and their expression level
dict_reprTrans (dict) : dict of genes and their
representative transcipt
intemediate_file (str) : txt file containing genes, transcript
and their expression level from the transkript_extractor function
Returns:
tsv file of representative trasncripts and their expression level
Raises:
None
"""
df_intermediate = repr.import_gtfSelection_to_df(intermediate_file)
df_geneTrans = transcripts_by_gene_inDf(df_intermediate)
df_exprTrans = tsv_or_csv_to_df(exprTrans)
df_reprTrans = dict_reprTrans_to_df(dict_reprTrans)
df_exprLevel_byGene = exprLevel_byGene(df_exprTrans, df_geneTrans)
df_match = match_byGene(df_reprTrans, df_exprLevel_byGene)
output = output_tsv(df_match)
return output
# run the programm
if __name__ == "__main__":
match_reprTranscript_expressionLevel()