list_from_bed.py

list_from_bed.py

import pandas as pd
import numpy as np
import os
import sys

bed_file = sys.argv[1]
name = bed_file.split('.')[0]

#import bed into dataframe
header = ['Chr', 'Start', 'End', 'Gene']
bed_df = pd.read_csv(bed_file, sep='\t', header=0, names=header, \
                     usecols=[0,1,2,3])
bed_df['Gene'] = bed_df['Gene'].apply(lambda x: x.split('.')[0])

list_df = pd.DataFrame(columns=['Chr', 'Start', 'End'])
#print bed_df['Gene'].unique()

f = open('%s.list'%name, 'w')
#print bed_df['Gene'].unique()


for gene in bed_df['Gene'].unique():

    #print gene

    temp_df  = bed_df[bed_df['Gene'] == gene]
    min_position = temp_df['Start'].min()
    max_position = temp_df['End'].max()

    if len(temp_df['Chr'].unique()) == 1:
        chromossome = temp_df['Chr'].iloc[0]
    else:
        print "Gene includes more than one chromossome, check the bed_file"
        break

    # new_inputs = [{'Chr' : chromossome, 'Start': min_position, 'End': max_position}]
    # new_df = pd.DataFrame(new_inputs)
    # list_df = list_df.append(new_df)
    string_to_write = str(chromossome) + ":" + str(min_position) + "-" \
                      + str(max_position) + "\n"
    f.write(string_to_write)
    print string_to_write