Mercurial > repos > glogobyte > viztool

comparison viz_ultra.py @ 13:41de387b3982 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author glogobyte
date Wed, 28 Oct 2020 07:32:14 +0000
parents 2c5723e2421a
children
comparison
equal deleted inserted replaced
12:3038dac9d61c 13:41de387b3982
1 import argparse
2 from functions import *
3 from viz_graphs import *
4 import sys
5 import pandas as pd
6 import matplotlib.pyplot as plt
7 import matplotlib.patches as mpatches
8 import matplotlib.font_manager as font_manager
9 import time
10 from multiprocessing import Process, Queue, Lock, Pool, Manager, Value
11
12
13 ##################################################################################################################################################################################################################
14
15 def top_diff(miRNA_info, number,flag,l):
16
17 Kind=[]
18
19 miRNA_info.sort(key = lambda x: abs(x[1]),reverse=True)
20 miRNA_info = miRNA_info[:number]
21 miRNA_info.sort(key = lambda x: x[0])
22
23 for x in miRNA_info:
24 if x[1] > 0:
25 Kind.append(True)
26 elif x[1] < 0:
27 Kind.append(False)
28 else:
29 Kind.append("Zero")
30
31 top_miRNA = {"Names": [x[0] for x in miRNA_info],
32 "Log2FC": [x[1] for x in miRNA_info],
33 "Kind": Kind};
34
35 df_miRNA = pd.DataFrame(data=top_miRNA)
36 df_miRNA = df_miRNA.sort_values(by=['Names'])
37 if df_miRNA.empty==False:
38 h1=df_miRNA.plot.barh(x= 'Names',y='Log2FC',color=df_miRNA.Kind.map({True: 'g', False: 'r', 'Zero':'k'}))
39 figure = plt.gcf() # get current figure
40 figure.set_size_inches(5, 12) # set figure's size manually to your full screen (32x18)
41 up_reg = mpatches.Patch(color='green', label='Upregulated')
42 down_reg = mpatches.Patch(color='red', label='Downregulated')
43 font = font_manager.FontProperties(weight='bold', style='normal')
44 l3 = plt.legend(handles=[up_reg,down_reg],bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0)
45 h1.set_ylabel(" ", fontsize=3, fontweight='bold')
46 h1.set_xlabel("Log2FC", fontsize=12, fontweight='bold')
47 plt.axvline(x=0, color="k")
48
49 plt.grid(axis='y', linewidth=0.2)
50 plt.grid(axis='x', linewidth=0.2)
51 if flag=='t':
52 plt.savefig('tem.png', bbox_inches='tight', dpi=300)
53 if flag=='nt':
54 plt.savefig('non.png', bbox_inches='tight', dpi=300)
55
56 ####################################################################################################################################################################################################################
57
58 def unique(sequence):
59 seen = set()
60 return [x for x in sequence if not (x in seen or seen.add(x))]
61
62 ###########################################################################################################################################################################################################################################################################
63
64 def top_scatter_non(matures,isoforms,non_temp,uni_names,number):
65
66 mat_names=[]
67 mat_log2fc=[]
68
69 iso_names=[]
70 iso_log2fc=[]
71
72 non_temp_names=[]
73 non_temp_log2fc=[]
74
75 count=0
76 for x in uni_names:
77 flag = False
78 if count<number:
79 for y in matures:
80 if x in y[0]:
81 mat_log2fc.append(y[1])
82 mat_names.append(x)
83 flag=True
84 for y in isoforms:
85 if x in y[0]:
86 iso_log2fc.append(y[1])
87 iso_names.append(x)
88 flag=True
89 for y in non_temp:
90 if x in y[0]:
91 non_temp_log2fc.append(y[1])
92 non_temp_names.append(x)
93 flag=True
94 if flag==True:
95 count+=1
96
97 mat_df = pd.DataFrame(dict(names=mat_names, log2fc=mat_log2fc))
98 iso_df = pd.DataFrame(dict(names=iso_names, log2fc=iso_log2fc))
99 non_df = pd.DataFrame(dict(names=non_temp_names, log2fc= non_temp_log2fc))
100
101 iso_df.sort_values(by=['names'])
102 mat_df.sort_values(by=['names'])
103 non_df.sort_values(by=['names'])
104
105 fig, ax = plt.subplots()
106
107 h3=ax.scatter(iso_df['log2fc'],iso_df['names'],edgecolors='k',linewidth=1, marker='o', c='red')
108 h1=ax.scatter(mat_df['log2fc'],mat_df['names'],edgecolors='k',linewidth=1, marker='o', c='green')
109 h2=ax.scatter(non_df['log2fc'],non_df['names'],edgecolors='k',linewidth=1, marker='o', c='blue')
110
111 l3 = plt.legend([h1,h2,h3],["Reference miRNA","Non-template","Template isomiRs"],bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0)
112 plt.axvline(x=0, color="k")
113 plt.grid(axis='y', linewidth=0.2)
114 plt.grid(axis='x', linewidth=0.2)
115 plt.xlabel("Log2FC", fontsize=12, fontweight='bold')
116 plt.yticks(rotation=0,ha="right", fontsize=10)
117 plt.xticks(rotation=0,ha="right", fontsize=10)
118 plt.tight_layout()
119 figure = plt.gcf() # get current figure
120 figure.set_size_inches(16, 12) # set figure's size manually to your full screen (32x18)
121 plt.savefig('a2.png', bbox_inches='tight', dpi=300)
122
123 #########################################################################################################################################################################################################################################
124 def top_scatter_tem(matures,isoforms,uni_names,number):
125
126 mat_names=[]
127 mat_log2fc=[]
128
129 iso_names=[]
130 iso_log2fc=[]
131
132 count=0
133 for x in uni_names:
134 flag = False
135 if count<number:
136 for y in matures:
137 if x in y[0]:
138 mat_log2fc.append(y[1])
139 mat_names.append(x)
140 flag=True
141 for y in isoforms:
142 if x in y[0]:
143 iso_log2fc.append(y[1])
144 iso_names.append(x)
145 flag=True
146 if flag==True:
147 count+=1
148
149 mat_df = pd.DataFrame(dict(names=mat_names, log2fc=mat_log2fc))
150 iso_df = pd.DataFrame(dict(names=iso_names, log2fc=iso_log2fc))
151
152 iso_df.sort_values(by=['names'])
153 mat_df.sort_values(by=['names'])
154
155 fig, ax = plt.subplots()
156
157 h3=ax.scatter(iso_df['log2fc'],iso_df['names'],edgecolors='k',linewidth=1, marker='o', c='red')
158 h1=ax.scatter(mat_df['log2fc'],mat_df['names'],edgecolors='k',linewidth=1, marker='o', c='green')
159
160 l3 = plt.legend([h1,h3],["Reference miRNA","Template isomiRs"],bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0)
161 plt.axvline(x=0, color="k")
162 plt.grid(axis='y', linewidth=0.2)
163 plt.grid(axis='x', linewidth=0.2)
164 plt.xlabel("Log2FC", fontsize=12, fontweight='bold')
165 plt.yticks(rotation=0,ha="right", fontsize=10)
166 plt.xticks(rotation=0,ha="right", fontsize=10)
167 plt.tight_layout()
168 figure = plt.gcf() # get current figure
169 figure.set_size_inches(16, 12) # set figure's size manually to your full screen (32x18)
170 plt.savefig('a2.png', bbox_inches='tight', dpi=300)
171
172
173 ##############################################################################################################################################################################################################################################
174 def preproccess(non_templated,matures,isoforms,log2fc,pval):
175
176 non_temp = [[x[0],float(x[1]),float(x[2])] for x in non_templated if abs(float(x[1]))>log2fc and float(x[2])<pval]
177 mat = [[x[0],float(x[1]),float(x[2])] for x in matures if abs(float(x[1]))>log2fc and float(x[2])<pval]
178 iso = [[x[0],float(x[1]),float(x[2])] for x in isoforms if abs(float(x[1]))>log2fc and float(x[2])<pval]
179 mat_iso = mat+iso
180
181 if not non_temp and not mat and not iso:
182 sys.exit("There aren't entries which meet these criteria")
183
184 mat.sort(key = lambda x: abs(float(x[1])),reverse=True)
185 iso.sort(key = lambda x: abs(float(x[1])),reverse=True)
186 non_temp.sort(key = lambda x: abs(float(x[1])),reverse=True)
187
188 all=mat+iso+non_temp
189 all.sort(key = lambda x: abs(float(x[1])), reverse=True)
190 names=[x[0].split("_")[0] for x in all]
191 uni_names=unique(names)
192
193 diff_non_templated = [[x[0],float(x[1]),float(x[2])] for x in non_templated if abs(float(x[1]))>1 and float(x[2])<pval and x[0].split("_")[0] in uni_names]
194 diff_matures = [[x[0],float(x[1]),float(x[2])] for x in matures if abs(float(x[1]))>1 and float(x[2])<pval and x[0].split("_")[0] in uni_names]
195 diff_isoforms = [[x[0],float(x[1]),float(x[2])] for x in isoforms if abs(float(x[1]))>1 and float(x[2])<pval and x[0].split("_")[0] in uni_names]
196
197 diff_matures.sort(key = lambda x: abs(float(x[1])),reverse=True)
198 diff_isoforms.sort(key = lambda x: abs(float(x[1])),reverse=True)
199 diff_non_templated.sort(key = lambda x: abs(float(x[1])),reverse=True)
200
201 return diff_matures,diff_isoforms,diff_non_templated,uni_names,non_temp,mat_iso
202
203 ##################################################################################################################################################################################################################################################################
204 starttime = time.time()
205
206 parser = argparse.ArgumentParser()
207 parser.add_argument("-in", "--input", help="choose type of analysis", action="store")
208 parser.add_argument("-p_value", "--pval", help="choose type of analysis", action="store")
209 parser.add_argument("-fc", "--log2fc", help="choose type of analysis", action="store")
210 parser.add_argument("-top", "--top_mirnas", help="choose type of analysis", action="store")
211 parser.add_argument("-tool_dir", "--tool_directory", help="tool directory path", action="store")
212 parser.add_argument("-statistic", "--stat", help="tool directory path", action="store")
213 parser.add_argument("-diff_tool", "--tool", help="tool directory path", action="store")
214
215 args = parser.parse_args()
216
217 l=Lock()
218 number = int(args.top_mirnas)
219 log2fc = float(args.log2fc)
220 pval = float(args.pval)
221
222 if args.tool=="2":
223
224 raw_EdgeR = read(args.input,0)
225 EdgeR = [x.rstrip("\n").split("\t") for x in raw_EdgeR]
226 del EdgeR[0]
227
228 if args.stat=="1":
229 non_templated = [[x[0],x[1],x[4]] for x in EdgeR if "__" in x[0] and x[1]!="NA" and x[4]!="NA"]
230 matures = [[x[0],x[1],x[4]] for x in EdgeR if 'chr' in x[0].split("_")[-1] and "__" not in x[0] and x[1]!="NA" and x[4]!="NA"]
231 isoforms = [[x[0],x[1],x[4]] for x in EdgeR if 'chr' not in x[0].split("_")[-1] and "__" not in x[0] and x[1]!="NA" and x[4]!="NA"]
232 else:
233 non_templated = [[x[0],x[1],x[5]] for x in EdgeR if "__" in x[0] and x[1]!="NA" and x[5]!="NA"]
234 matures = [[x[0],x[1],x[5]] for x in EdgeR if 'chr' in x[0].split("_")[-1] and "__" not in x[0] and x[1]!="NA" and x[5]!="NA"]
235 isoforms = [[x[0],x[1],x[5]] for x in EdgeR if 'chr' not in x[0].split("_")[-1] and "__" not in x[0] and x[1]!="NA" and x[5]!="NA"]
236
237 if args.tool=="1":
238
239 raw_Deseq = read(args.input,0)
240 Deseq = [x.rstrip("\n").split("\t") for x in raw_Deseq]
241
242 if args.stat=="1":
243 non_templated = [[x[0],x[2],x[5]] for x in Deseq if "__" in x[0] and x[2]!="NA" and x[5]!="NA"]
244 matures = [[x[0],x[2],x[5]] for x in Deseq if 'chr' in x[0].split("_")[-1] and "__" not in x[0] and x[2]!="NA" and x[5]!="NA"]
245 isoforms = [[x[0],x[2],x[5]] for x in Deseq if 'chr' not in x[0].split("_")[-1] and "__" not in x[0] and x[2]!="NA" and x[5]!="NA"]
246 else:
247 non_templated = [[x[0],x[2],x[6]] for x in Deseq if "__" in x[0] and x[2]!="NA" and x[6]!="NA"]
248 matures = [[x[0],x[2],x[6]] for x in Deseq if 'chr' in x[0].split("_")[-1] and "__" not in x[0] and x[2]!="NA" and x[6]!="NA"]
249 isoforms = [[x[0],x[2],x[6]] for x in Deseq if 'chr' not in x[0].split("_")[-1] and "__" not in x[0] and x[2]!="NA" and x[6]!="NA"]
250
251
252 diff_matures,diff_isoforms,diff_non_templated,names,non_temp,mat_iso = preproccess(non_templated,matures,isoforms,log2fc,pval)
253
254 if non_templated!=[]:
255 analysis="2"
256 p=[Process(target=top_diff,args=(non_temp,number,"nt",l))]
257 p.extend([Process(target=top_diff,args=(mat_iso,number,"t",l))])
258 p.extend([Process(target=top_scatter_non,args=(diff_matures,diff_isoforms,diff_non_templated,names,number))])
259
260 else:
261 analysis="1"
262 p=[Process(target=top_diff,args=(mat_iso,number,"t"))]
263 p.extend([Process(target=top_scatter_tem,args=(diff_matures,diff_isoforms,names,number))])
264
265 [x.start() for x in p]
266 [x.join() for x in p]
267
268 pdf_after_DE(analysis,args.top_mirnas)
269
270 print('That took {} seconds'.format(time.time() - starttime))
271