Mercurial > repos > chmaramis > testirprofiler
diff comp_clono_CDR3.py @ 12:cdf95051bc55 draft default tip
Uploaded 2 tools
| author | chmaramis |
|---|---|
| date | Sun, 18 Mar 2018 07:11:06 -0400 |
| parents | |
| children |
line wrap: on
line diff
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/comp_clono_CDR3.py Sun Mar 18 07:11:06 2018 -0400 @@ -0,0 +1,78 @@ +# -*- coding: utf-8 -*- +""" +Created on Thu Jun 19 17:33:34 2014 + +@author: chmaramis +""" + +from __future__ import division +import numpy as np +from pandas import * +import functools as ft +import sys +import time + +frm = lambda x,y: '{r}/{l}'.format(r=x,l=y) + + +def cdr3Computation(inp_name, out1, t10n, fname): + + frame = DataFrame() + tp = read_csv(inp_name, iterator=True, chunksize=5000,sep='\t', index_col=0 ) + frame = concat([chunk for chunk in tp]) + + grouped = frame.groupby(['AA JUNCTION']) + x=grouped.size() + x1=DataFrame(x.index, columns=['AA JUNCTION']) + x1['Reads']=x.values + total = sum(x1['Reads']) + #x1['Reads/Total'] = ['{r}/{l}'.format(r=pr , l = total) for pr in x1['Reads']] + x1['Reads/Total'] = x1['Reads'].map(ft.partial(frm, y=total)) + x1['Frequency %'] = (100*x1['Reads']/total).map('{:.4f}'.format) + + final = x1.sort_values(by = ['Reads'] , ascending = False) + + final.index=range(1,len(final)+1) + final.to_csv(out1 , sep = '\t') + + numofclono = len(final) + clust = len(final[final['Reads'] > 1]) + sing = len (final[final['Reads'] == 1]) + top10 = final[['AA JUNCTION','Frequency %']].head(10) + top10.to_csv(t10n , sep = '\t') + + summary = [[str(top10['AA JUNCTION'].values[0])]] + summary.append([top10['Frequency %'].values[0]]) + summary.append([numofclono]) + summary.append([clust,'{:.4f}'.format(100*clust/numofclono)]) + summary.append([sing,'{:.4f}'.format(100*sing/numofclono)]) + + ind = ['Dominant Clonotype (CDR3)', 'Frequency', 'Number of Clonotypes' , 'Expanding Clonotypes','Singletons'] + spl = fname.split('_') + col = [spl[0],'%'] + + frsum = DataFrame(summary,index = ind, columns = col) + + return frsum + +if __name__ == '__main__': + + start=time.time() + + # Parse input arguments + inp_name = sys.argv[1] + out1 = sys.argv[2] + t10n = sys.argv[3] + sname = sys.argv[4] + fname = sys.argv[5] + + # Execute basic function + frsum = cdr3Computation(inp_name,out1,t10n,fname) + + # Save output to CSV files + if not frsum.empty: + frsum.to_csv(sname, sep = '\t') + + # Print execution time + stop=time.time() + print('Runtime:' + str(stop-start))
