tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

tidy code and renamed kd.py to kd_df.py

Browse source
This commit is contained in:
Tanushree Tunstall 2020-03-26 15:43:13 +00:00
parent 73e0029b65
commit a074d29f6e
6 changed files with 156 additions and 194 deletions
Download patch file Download diff file Expand all files Collapse all files

28
meta_data_analysis/mut_electrostatic_changes.py
View file

@ -10,15 +10,17 @@ Created on Tue Aug 6 12:56:03 2019
#=======================================================================
# TASK: calculate how many mutations result in
# electrostatic changes wrt wt
# Input: mcsm and AF_OR file
# Output: mut_elec_changes_results.txt
#=======================================================================
#%% load libraries
import os, sys
import pandas as pd
#import numpy as np
#%% specify homedir as python doesn't recognise tilde
#=======================================================================
#%% specify homedir and curr dir
homedir = os.path.expanduser('~')
# set working dir
@ -26,7 +28,7 @@ os.getcwd()
os.chdir(homedir + '/git/LSHTM_analysis/meta_data_analysis')
os.getcwd()
#=======================================================================
#%% variable assignment: input and output paths & filenames
#%% variable assignment: input and output
drug = 'pyrazinamide'
gene = 'pncA'
gene_match = gene + '_p.'
@ -59,7 +61,6 @@ print('Output filename: ', out_filename
#%% end of variable assignment for input and output files
#=======================================================================
#%% Read input files
#in_filename = gene.lower() + '_meta_data_with_AFandOR.csv'
print('Reading input file (merged file):', infile)
comb_df = pd.read_csv(infile, sep = ',')
@ -136,17 +137,11 @@ print('Total number of electrostatic changes resulting from Mutation is (%):', e
no_change_muts = ap_df[ap_df['aa_calcprop'].isin(['neg->neg','non-polar->non-polar','polar->polar', 'pos->pos']) == True]
no_change_muts.mut_count.sum()
###########
# Step 5: output from console
###########
#%% output from console
#sys.stdout = open(file, 'w')
sys.stdout = open(outfile, 'w')
#print(no_change_muts, '\n',
# all_prop_change)
print('======================\n'
,'Unchanged muts'
,'\n=====================\n'
@ -156,16 +151,11 @@ print('======================\n'
, '\n============================\n'
, all_prop_change)
#print('======================================================================')
#print('Total number of electrostatic changes resulting from Mutation is (%):', elec_changes)
#print('Total no. of muts: ', total_muts)
#print('Total no. of changed muts: ', all_prop_change.mut_count.sum())
#print('Total no. of unchanged muts: ', no_change_muts.mut_count.sum() )
#print('=======================================================================')
print('========================================================================'
, '\nTotal number of electrostatic changes resulting from Mtation is (%):', elec_changes
, '\nTotal no. of muts: ', total_muts
, '\nTotal no. of changed muts: ', all_prop_change.mut_count.sum()
, '\nTotal no. of unchanged muts: ', no_change_muts.mut_count.sum()
, '\n=========================================================================')
#%% end of script
#=======================================================================