src/db_utils/frequency_scripts/create_n_fill_exon_frequency_file.py :...

src/db_utils/frequency_scripts/create_n_fill_exon_frequency_file.py : modification of compute_gene_orf_data to compute codon/aa/properties frequencies at gene level

src/db_utils/frequency_scripts/create_n_fill_exon_frequency_file.py

@@ -119,14 +119,20 @@ def compute_gene_orf_data(df: pd.DataFrame) -> pd.DataFrame:
     :return:
     """
     df['id_gene'] = df['id_exon'].str.replace(r'_\d+', '')
-    gene_len = df[['id_gene', 'length']].groupby('id_gene').sum().reset_index()
+    df_exon = df.loc[(df['ft_type'] == "aa") & (df['ft'] == 'A'), :]
+    res = df_exon.groupby('id_exon').count()['length'].values
+    if np.max(res) != np.min(res) != 1:
+        msg = "On exon must be present only once !"
+        logging.exception(msg)
+        raise ValueError(msg)
+    gene_len = df_exon[['id_gene', 'length']].groupby('id_gene').sum()\
+        .reset_index()
     df = pd.merge(df, gene_len, how='left', on="id_gene",
                   suffixes=['_exon', '_gene'])
     df['new_freq'] = df['frequency'] * (df['length_exon'] / df['length_gene'])
     df_final = df[['id_gene', 'ft_type', 'ft', 'new_freq']].\
         groupby(['id_gene', 'ft_type', 'ft']).sum().reset_index()
     df_final.rename(columns={"new_freq": 'frequency'}, inplace=True)
-    df_final['frequency'] = round(df_final['frequency'], 5)
     return df_final
 
 
@@ -174,6 +180,8 @@ def create_or_load_freq_table(ps: int) -> Tuple[pd.DataFrame, pd.DataFrame]:
     else:
         df_gene = pd.read_csv(Config_freq.output_gene_file, sep="\t")
     HG19 = None
+    df_exon['frequency'] = round(df_exon['frequency'], 5)
+    df_gene['frequency'] = round(df_gene['frequency'], 5)
     return df_exon, df_gene