Mercurial > repos > glogobyte > viztool
comparison viz_ultra.py @ 15:6db3bd727fde draft
Uploaded
| author | glogobyte |
|---|---|
| date | Wed, 28 Oct 2020 07:34:56 +0000 |
| parents | |
| children |
comparison
equal
deleted
inserted
replaced
| 14:e51ebc767701 | 15:6db3bd727fde |
|---|---|
| 1 import argparse | |
| 2 from viz_graphs import * | |
| 3 import sys | |
| 4 import pandas as pd | |
| 5 import matplotlib.pyplot as plt | |
| 6 import matplotlib.patches as mpatches | |
| 7 import matplotlib.font_manager as font_manager | |
| 8 import time | |
| 9 from multiprocessing import Process, Queue, Lock, Pool, Manager, Value | |
| 10 | |
| 11 | |
| 12 ################################################################################################################################################################################################################## | |
| 13 | |
| 14 def top_diff(miRNA_info, number,flag,l): | |
| 15 | |
| 16 Kind=[] | |
| 17 | |
| 18 miRNA_info.sort(key = lambda x: abs(x[1]),reverse=True) | |
| 19 miRNA_info = miRNA_info[:number] | |
| 20 miRNA_info.sort(key = lambda x: x[0]) | |
| 21 | |
| 22 for x in miRNA_info: | |
| 23 if x[1] > 0: | |
| 24 Kind.append(True) | |
| 25 elif x[1] < 0: | |
| 26 Kind.append(False) | |
| 27 else: | |
| 28 Kind.append("Zero") | |
| 29 | |
| 30 top_miRNA = {"Names": [x[0] for x in miRNA_info], | |
| 31 "Log2FC": [x[1] for x in miRNA_info], | |
| 32 "Kind": Kind}; | |
| 33 | |
| 34 df_miRNA = pd.DataFrame(data=top_miRNA) | |
| 35 df_miRNA = df_miRNA.sort_values(by=['Names']) | |
| 36 if df_miRNA.empty==False: | |
| 37 h1=df_miRNA.plot.barh(x= 'Names',y='Log2FC',color=df_miRNA.Kind.map({True: 'g', False: 'r', 'Zero':'k'})) | |
| 38 figure = plt.gcf() # get current figure | |
| 39 figure.set_size_inches(5, 12) # set figure's size manually to your full screen (32x18) | |
| 40 up_reg = mpatches.Patch(color='green', label='Upregulated') | |
| 41 down_reg = mpatches.Patch(color='red', label='Downregulated') | |
| 42 font = font_manager.FontProperties(weight='bold', style='normal') | |
| 43 l3 = plt.legend(handles=[up_reg,down_reg],bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0) | |
| 44 h1.set_ylabel(" ", fontsize=3, fontweight='bold') | |
| 45 h1.set_xlabel("Log2FC", fontsize=12, fontweight='bold') | |
| 46 plt.axvline(x=0, color="k") | |
| 47 | |
| 48 plt.grid(axis='y', linewidth=0.2) | |
| 49 plt.grid(axis='x', linewidth=0.2) | |
| 50 if flag=='t': | |
| 51 plt.savefig('tem.png', bbox_inches='tight', dpi=300) | |
| 52 if flag=='nt': | |
| 53 plt.savefig('non.png', bbox_inches='tight', dpi=300) | |
| 54 | |
| 55 #################################################################################################################################################################################################################### | |
| 56 | |
| 57 def unique(sequence): | |
| 58 seen = set() | |
| 59 return [x for x in sequence if not (x in seen or seen.add(x))] | |
| 60 | |
| 61 ########################################################################################################################################################################################################################################################################### | |
| 62 | |
| 63 def top_scatter_non(matures,isoforms,non_temp,uni_names,number): | |
| 64 | |
| 65 mat_names=[] | |
| 66 mat_log2fc=[] | |
| 67 | |
| 68 iso_names=[] | |
| 69 iso_log2fc=[] | |
| 70 | |
| 71 non_temp_names=[] | |
| 72 non_temp_log2fc=[] | |
| 73 | |
| 74 count=0 | |
| 75 for x in uni_names: | |
| 76 flag = False | |
| 77 if count<number: | |
| 78 for y in matures: | |
| 79 if x in y[0]: | |
| 80 mat_log2fc.append(y[1]) | |
| 81 mat_names.append(x) | |
| 82 flag=True | |
| 83 for y in isoforms: | |
| 84 if x in y[0]: | |
| 85 iso_log2fc.append(y[1]) | |
| 86 iso_names.append(x) | |
| 87 flag=True | |
| 88 for y in non_temp: | |
| 89 if x in y[0]: | |
| 90 non_temp_log2fc.append(y[1]) | |
| 91 non_temp_names.append(x) | |
| 92 flag=True | |
| 93 if flag==True: | |
| 94 count+=1 | |
| 95 | |
| 96 mat_df = pd.DataFrame(dict(names=mat_names, log2fc=mat_log2fc)) | |
| 97 iso_df = pd.DataFrame(dict(names=iso_names, log2fc=iso_log2fc)) | |
| 98 non_df = pd.DataFrame(dict(names=non_temp_names, log2fc= non_temp_log2fc)) | |
| 99 | |
| 100 iso_df.sort_values(by=['names']) | |
| 101 mat_df.sort_values(by=['names']) | |
| 102 non_df.sort_values(by=['names']) | |
| 103 | |
| 104 fig, ax = plt.subplots() | |
| 105 | |
| 106 h3=ax.scatter(iso_df['log2fc'],iso_df['names'],edgecolors='k',linewidth=1, marker='o', c='red') | |
| 107 h1=ax.scatter(mat_df['log2fc'],mat_df['names'],edgecolors='k',linewidth=1, marker='o', c='green') | |
| 108 h2=ax.scatter(non_df['log2fc'],non_df['names'],edgecolors='k',linewidth=1, marker='o', c='blue') | |
| 109 | |
| 110 l3 = plt.legend([h1,h2,h3],["Reference miRNA","Non-template","Template isomiRs"],bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0) | |
| 111 plt.axvline(x=0, color="k") | |
| 112 plt.grid(axis='y', linewidth=0.2) | |
| 113 plt.grid(axis='x', linewidth=0.2) | |
| 114 plt.xlabel("Log2FC", fontsize=12, fontweight='bold') | |
| 115 plt.yticks(rotation=0,ha="right", fontsize=10) | |
| 116 plt.xticks(rotation=0,ha="right", fontsize=10) | |
| 117 plt.tight_layout() | |
| 118 figure = plt.gcf() # get current figure | |
| 119 figure.set_size_inches(16, 12) # set figure's size manually to your full screen (32x18) | |
| 120 plt.savefig('a2.png', bbox_inches='tight', dpi=300) | |
| 121 | |
| 122 ######################################################################################################################################################################################################################################### | |
| 123 def top_scatter_tem(matures,isoforms,uni_names,number): | |
| 124 | |
| 125 mat_names=[] | |
| 126 mat_log2fc=[] | |
| 127 | |
| 128 iso_names=[] | |
| 129 iso_log2fc=[] | |
| 130 | |
| 131 count=0 | |
| 132 for x in uni_names: | |
| 133 flag = False | |
| 134 if count<number: | |
| 135 for y in matures: | |
| 136 if x in y[0]: | |
| 137 mat_log2fc.append(y[1]) | |
| 138 mat_names.append(x) | |
| 139 flag=True | |
| 140 for y in isoforms: | |
| 141 if x in y[0]: | |
| 142 iso_log2fc.append(y[1]) | |
| 143 iso_names.append(x) | |
| 144 flag=True | |
| 145 if flag==True: | |
| 146 count+=1 | |
| 147 | |
| 148 mat_df = pd.DataFrame(dict(names=mat_names, log2fc=mat_log2fc)) | |
| 149 iso_df = pd.DataFrame(dict(names=iso_names, log2fc=iso_log2fc)) | |
| 150 | |
| 151 iso_df.sort_values(by=['names']) | |
| 152 mat_df.sort_values(by=['names']) | |
| 153 | |
| 154 fig, ax = plt.subplots() | |
| 155 | |
| 156 h3=ax.scatter(iso_df['log2fc'],iso_df['names'],edgecolors='k',linewidth=1, marker='o', c='red') | |
| 157 h1=ax.scatter(mat_df['log2fc'],mat_df['names'],edgecolors='k',linewidth=1, marker='o', c='green') | |
| 158 | |
| 159 l3 = plt.legend([h1,h3],["Reference miRNA","Template isomiRs"],bbox_to_anchor=(1.04,0.5), loc="center left", borderaxespad=0) | |
| 160 plt.axvline(x=0, color="k") | |
| 161 plt.grid(axis='y', linewidth=0.2) | |
| 162 plt.grid(axis='x', linewidth=0.2) | |
| 163 plt.xlabel("Log2FC", fontsize=12, fontweight='bold') | |
| 164 plt.yticks(rotation=0,ha="right", fontsize=10) | |
| 165 plt.xticks(rotation=0,ha="right", fontsize=10) | |
| 166 plt.tight_layout() | |
| 167 figure = plt.gcf() # get current figure | |
| 168 figure.set_size_inches(16, 12) # set figure's size manually to your full screen (32x18) | |
| 169 plt.savefig('a2.png', bbox_inches='tight', dpi=300) | |
| 170 | |
| 171 | |
| 172 ############################################################################################################################################################################################################################################## | |
| 173 def preproccess(non_templated,matures,isoforms,log2fc,pval): | |
| 174 | |
| 175 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] | |
| 176 mat = [[x[0],float(x[1]),float(x[2])] for x in matures if abs(float(x[1]))>log2fc and float(x[2])<pval] | |
| 177 iso = [[x[0],float(x[1]),float(x[2])] for x in isoforms if abs(float(x[1]))>log2fc and float(x[2])<pval] | |
| 178 mat_iso = mat+iso | |
| 179 | |
| 180 if not non_temp and not mat and not iso: | |
| 181 sys.exit("There aren't entries which meet these criteria") | |
| 182 | |
| 183 mat.sort(key = lambda x: abs(float(x[1])),reverse=True) | |
| 184 iso.sort(key = lambda x: abs(float(x[1])),reverse=True) | |
| 185 non_temp.sort(key = lambda x: abs(float(x[1])),reverse=True) | |
| 186 | |
| 187 all=mat+iso+non_temp | |
| 188 all.sort(key = lambda x: abs(float(x[1])), reverse=True) | |
| 189 names=[x[0].split("_")[0] for x in all] | |
| 190 uni_names=unique(names) | |
| 191 | |
| 192 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] | |
| 193 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] | |
| 194 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] | |
| 195 | |
| 196 diff_matures.sort(key = lambda x: abs(float(x[1])),reverse=True) | |
| 197 diff_isoforms.sort(key = lambda x: abs(float(x[1])),reverse=True) | |
| 198 diff_non_templated.sort(key = lambda x: abs(float(x[1])),reverse=True) | |
| 199 | |
| 200 return diff_matures,diff_isoforms,diff_non_templated,uni_names,non_temp,mat_iso | |
| 201 | |
| 202 ################################################################################################################################################################################################################################################################## | |
| 203 starttime = time.time() | |
| 204 | |
| 205 parser = argparse.ArgumentParser() | |
| 206 parser.add_argument("-in", "--input", help="choose type of analysis", action="store") | |
| 207 parser.add_argument("-p_value", "--pval", help="choose type of analysis", action="store") | |
| 208 parser.add_argument("-fc", "--log2fc", help="choose type of analysis", action="store") | |
| 209 parser.add_argument("-top", "--top_mirnas", help="choose type of analysis", action="store") | |
| 210 parser.add_argument("-tool_dir", "--tool_directory", help="tool directory path", action="store") | |
| 211 parser.add_argument("-statistic", "--stat", help="tool directory path", action="store") | |
| 212 parser.add_argument("-diff_tool", "--tool", help="tool directory path", action="store") | |
| 213 | |
| 214 args = parser.parse_args() | |
| 215 | |
| 216 l=Lock() | |
| 217 number = int(args.top_mirnas) | |
| 218 log2fc = float(args.log2fc) | |
| 219 pval = float(args.pval) | |
| 220 | |
| 221 if args.tool=="2": | |
| 222 | |
| 223 raw_EdgeR = read(args.input,0) | |
| 224 EdgeR = [x.rstrip("\n").split("\t") for x in raw_EdgeR] | |
| 225 del EdgeR[0] | |
| 226 | |
| 227 if args.stat=="1": | |
| 228 non_templated = [[x[0],x[1],x[4]] for x in EdgeR if "__" in x[0] and x[1]!="NA" and x[4]!="NA"] | |
| 229 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"] | |
| 230 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"] | |
| 231 else: | |
| 232 non_templated = [[x[0],x[1],x[5]] for x in EdgeR if "__" in x[0] and x[1]!="NA" and x[5]!="NA"] | |
| 233 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"] | |
| 234 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"] | |
| 235 | |
| 236 if args.tool=="1": | |
| 237 | |
| 238 raw_Deseq = read(args.input,0) | |
| 239 Deseq = [x.rstrip("\n").split("\t") for x in raw_Deseq] | |
| 240 | |
| 241 if args.stat=="1": | |
| 242 non_templated = [[x[0],x[2],x[5]] for x in Deseq if "__" in x[0] and x[2]!="NA" and x[5]!="NA"] | |
| 243 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"] | |
| 244 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"] | |
| 245 else: | |
| 246 non_templated = [[x[0],x[2],x[6]] for x in Deseq if "__" in x[0] and x[2]!="NA" and x[6]!="NA"] | |
| 247 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"] | |
| 248 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"] | |
| 249 | |
| 250 | |
| 251 diff_matures,diff_isoforms,diff_non_templated,names,non_temp,mat_iso = preproccess(non_templated,matures,isoforms,log2fc,pval) | |
| 252 | |
| 253 if non_templated!=[]: | |
| 254 analysis="2" | |
| 255 p=[Process(target=top_diff,args=(non_temp,number,"nt",l))] | |
| 256 p.extend([Process(target=top_diff,args=(mat_iso,number,"t",l))]) | |
| 257 p.extend([Process(target=top_scatter_non,args=(diff_matures,diff_isoforms,diff_non_templated,names,number))]) | |
| 258 | |
| 259 else: | |
| 260 analysis="1" | |
| 261 p=[Process(target=top_diff,args=(mat_iso,number,"t"))] | |
| 262 p.extend([Process(target=top_scatter_tem,args=(diff_matures,diff_isoforms,names,number))]) | |
| 263 | |
| 264 [x.start() for x in p] | |
| 265 [x.join() for x in p] | |
| 266 | |
| 267 pdf_after_DE(analysis,args.top_mirnas) | |
| 268 | |
| 269 print('That took {} seconds'.format(time.time() - starttime)) | |
| 270 |