|
4
|
1 # A translation of aggregate.pl into python! For analysis of Tn-Seq.
|
|
|
2 # This script requires BioPython just like calc_fitness.py, so you need it installed along with its dependencies if you want to run these scripts on your own.
|
|
|
3 # How to install BioPython and a list of its dependencies can be found here: http://biopython.org/DIST/docs/install/Installation.html
|
|
|
4
|
|
|
5
|
|
|
6
|
|
|
7
|
|
|
8
|
|
|
9
|
|
|
10
|
|
|
11
|
|
|
12
|
|
|
13
|
|
|
14 ##### ARGUMENTS #####
|
|
|
15
|
|
|
16 # Prints basic instructions on / options for using this code; called when the user forgets to enter an output file or fitness file(s)
|
|
|
17
|
|
|
18 def print_usage():
|
|
|
19 print "Aggregate.py's usage is as follows:" + "\n\n"
|
|
|
20 print "\033[1m" + "Required" + "\033[0m" + "\n"
|
|
|
21 print "-o" + "\t\t" + "Output file for aggregated data." + "\n"
|
|
|
22 print "\n"
|
|
|
23 print "\033[1m" + "Optional" + "\033[0m" + "\n"
|
|
|
24 print "-c" + "\t\t" + "Check for missing genes in the data set - provide a reference genome in genbank format. Missing genes will be sent to stdout." + "\n"
|
|
|
25 print "-m" + "\t\t" + "Place a mark in an extra column for this set of genes. Provide a file with a list of genes seperated by newlines." + "\n"
|
|
|
26 print "-x" + "\t\t" + "Cutoff: Don't include fitness scores with average counts (c1+c2)/2 < x (default: 0)" + "\n"
|
|
|
27 print "-b" + "\t\t" + "Blanks: Exclude -b % of blank fitness scores (scores where c2 = 0) (default: 0 = 0%)" + "\n"
|
|
|
28 print "-w" + "\t\t" + "Use weighted algorithm to calculate averages, variance, sd, se" + "\n"
|
|
|
29 print "-l" + "\t\t" + "Weight ceiling: maximum value to use as a weight (default: 999,999)" + "\n"
|
|
|
30 print "\n"
|
|
|
31 print "All remainder arguements will be treated as fitness files (those files created by calc_fitness.py)" + "\n"
|
|
|
32 print "\n"
|
|
|
33
|
|
|
34 # Turns the arguments entered in the command line into variables so that they can actually be called.
|
|
|
35
|
|
|
36 import argparse
|
|
|
37 parser = argparse.ArgumentParser()
|
|
|
38 parser.add_argument("-o", action="store", dest="summary")
|
|
|
39 parser.add_argument("-c", action="store", dest="find_missing")
|
|
|
40 parser.add_argument("-m", action="store", dest="marked")
|
|
|
41 parser.add_argument("-x", action="store", dest="cutoff")
|
|
|
42 parser.add_argument("-b", action="store", dest="blank_pc")
|
|
|
43 parser.add_argument("-w", action="store", dest="weighted")
|
|
|
44 parser.add_argument("-l", action="store", dest="weight_ceiling")
|
|
|
45 parser.add_argument("fitnessfiles", nargs=argparse.REMAINDER)
|
|
|
46
|
|
|
47 #Shortens the names of those variables created from the arguments - from parser.parse_args().ref_genome to arguments.ref_genome, for example.
|
|
|
48
|
|
|
49 arguments = parser.parse_args()
|
|
|
50
|
|
|
51 #Checks that the required arguments have been entered; if not informs the user they need to provide a name for the output file / the fitness file(s) and print's aggregate.py's usage
|
|
|
52
|
|
|
53 if not arguments.summary:
|
|
|
54 print "\n" + "You are missing a value for the -o flag. "
|
|
|
55 print_usage()
|
|
|
56 quit()
|
|
|
57
|
|
|
58 if not arguments.fitnessfiles:
|
|
|
59 print "\n" + "You are missing fitness file(s); these should be entered immediately after all the flags. "
|
|
|
60 print_usage()
|
|
|
61 quit()
|
|
|
62
|
|
|
63 #Sets the maximum weight of a fitness value, if that wasn't specified in the command line.
|
|
|
64
|
|
|
65 if (not arguments.weight_ceiling):
|
|
|
66 arguments.max_weight = 999999
|
|
|
67
|
|
|
68 #Sets the cutoff to a default value of 0 if it wasn't specified in the command line.
|
|
|
69 #Cutoff exists to discard positions with a low number of counted transcripts, because their fitness may not be as accurate - for the same reasoning that studies with low sample sizes can be innacurate.
|
|
|
70
|
|
|
71 if (not arguments.cutoff):
|
|
|
72 arguments.cutoff = 0
|
|
|
73
|
|
|
74 #Sets the % of blank fitness values to exclude to a default value of 0 if it wasn't specified in the command line. This can be found in the 2nd output of calc_fitness.
|
|
|
75
|
|
|
76 if (not arguments.blank_pc):
|
|
|
77 arguments.blank_pc = 0
|
|
|
78
|
|
|
79 # gets blank_pc from the output file of calc_fit / consol
|
|
|
80
|
|
|
81 if arguments.blank_pc:
|
|
|
82 with open(arguments.blank_pc) as file:
|
|
|
83 blank_pc = file.read().splitlines()
|
|
|
84 arguments.blank_pc = float(blank_pc[1].split()[1])
|
|
|
85
|
|
|
86
|
|
|
87
|
|
|
88
|
|
|
89
|
|
|
90
|
|
|
91
|
|
|
92 ##### SUBROUTINES #####
|
|
|
93
|
|
|
94 #A subroutine that calculates the average, variance, standard deviation (sd), and standard error (se) of a group of scores; for use when aggregating scores by gene later on
|
|
|
95
|
|
|
96 import math
|
|
|
97 def average(scores):
|
|
|
98 sum = 0
|
|
|
99 num = 0
|
|
|
100
|
|
|
101 #Finds the average of the scores
|
|
|
102
|
|
|
103 for i in scores:
|
|
|
104 sum += i
|
|
|
105 num += 1
|
|
|
106 average = sum/num
|
|
|
107
|
|
|
108 #Finds the variance of the scores
|
|
|
109
|
|
|
110 xminusxbars = 0
|
|
|
111 for i in scores:
|
|
|
112 xminusxbars += (i - average)**2
|
|
|
113 variance = xminusxbars/(num-1)
|
|
|
114
|
|
|
115 #Finds the standard deviation and standard error of the scores; then the average / variance / standard deviation / standard error are returned
|
|
|
116
|
|
|
117 sd = math.sqrt(variance)
|
|
|
118 se = sd / math.sqrt(num)
|
|
|
119 return (average, variance, sd, se)
|
|
|
120
|
|
|
121 #A subroutine that calculates the weighted average, variance, standard deviation (sd), and standard error (se) of a group of scores; the weights come from the number of reads each insertion location has
|
|
|
122 #For use when aggregating scores by gene later on, if the weighted argument is called
|
|
|
123
|
|
|
124 def weighted_average(scores,weights):
|
|
|
125 sum = 0
|
|
|
126 weighted_average = 0
|
|
|
127 weighted_variance = 0
|
|
|
128 top = 0
|
|
|
129 bottom = 0
|
|
|
130 i = 0
|
|
|
131
|
|
|
132 #Finds weighted average of the scores
|
|
|
133
|
|
|
134 while i < len(weights):
|
|
|
135 if not scores[i]:
|
|
|
136 scores[i] = 0.0
|
|
|
137 top += float(weights[i])*float(scores[i])
|
|
|
138 bottom += float(weights[i])
|
|
|
139 i += 1
|
|
|
140 if bottom == 0:
|
|
|
141 return 0
|
|
|
142 weighted_average = top/bottom
|
|
|
143
|
|
|
144 #Finds weighted variance of the scores
|
|
|
145
|
|
|
146 top = 0
|
|
|
147 bottom = 0
|
|
|
148 i = 0
|
|
|
149 while i < len(weights):
|
|
|
150 top += float(weights[i]) * (float(scores[i]) - weighted_average)**2
|
|
|
151 bottom += float(weights[i])
|
|
|
152 i += 1
|
|
|
153 weighted_variance = top/bottom
|
|
|
154
|
|
|
155 #Finds weighted standard deviation and standard error of the scores; then the weighted average / variance / standard deviation / standard error are returned
|
|
|
156
|
|
|
157 weighted_stdev = math.sqrt(weighted_variance)
|
|
|
158 weighted_stder = weighted_stdev/math.sqrt(len(scores))
|
|
|
159 return (weighted_average, weighted_variance, weighted_stdev, weighted_stder)
|
|
|
160
|
|
|
161
|
|
|
162
|
|
|
163
|
|
|
164
|
|
|
165
|
|
|
166
|
|
|
167
|
|
|
168
|
|
|
169
|
|
|
170 ##### AGGREGATION / CALCULATIONS #####
|
|
|
171
|
|
|
172 #Reads the genes which should be marked in the final aggregate file into an array
|
|
|
173
|
|
|
174 import os.path
|
|
|
175 if arguments.marked:
|
|
|
176 with open(arguments.marked) as file:
|
|
|
177 marked_set = file.read().splitlines()
|
|
|
178
|
|
|
179 #Creates a dictionary of dictionaries to contain a summary of all genes and their fitness values
|
|
|
180 #Each gene is its own dictionary with w and s as keys for the various fitnesses and weights of the insertion locations within those genes respectively
|
|
|
181 #The fitness values and weights match up, so that the weight of gene_summary[locus]["w"][2] would be gene_summary[locus]["s"][2]
|
|
|
182
|
|
|
183 import csv
|
|
|
184 gene_summary = {}
|
|
|
185 for eachfile in arguments.fitnessfiles:
|
|
|
186 with open(eachfile) as csvfile:
|
|
|
187 lines = csv.reader(csvfile)
|
|
|
188 for line in lines:
|
|
|
189 locus = line[9]
|
|
|
190 w = line[12]
|
|
|
191 if w == 'nW':
|
|
|
192 continue
|
|
|
193 if not w:
|
|
|
194 w == 0
|
|
|
195 c1 = float(line[2])
|
|
|
196 c2 = float(line[3])
|
|
|
197 avg = (c1+c2)/2
|
|
|
198 if avg < float(arguments.cutoff):
|
|
|
199 continue
|
|
|
200 if avg > float(arguments.weight_ceiling):
|
|
|
201 avg = arguments.weight_ceiling
|
|
|
202 if locus not in gene_summary:
|
|
|
203 gene_summary[locus] = {"w" : [], "s": []}
|
|
|
204 gene_summary[locus]["w"].append(w)
|
|
|
205 gene_summary[locus]["s"].append(avg)
|
|
|
206
|
|
|
207 #If finding any missing gene loci is requested in the arguments, starts out by loading all the known features from a genbank file
|
|
|
208
|
|
|
209 from Bio import SeqIO
|
|
|
210 if (arguments.find_missing):
|
|
|
211 output = [["locus","mean","var","sd","se","gene","Total","Blank","Not Blank","Blank Removed","M\n"]]
|
|
|
212 handle = open(arguments.find_missing, "rU")
|
|
|
213 for record in SeqIO.parse(handle, "genbank"):
|
|
|
214 refname = record.id
|
|
|
215 features = record.features
|
|
|
216 handle.close()
|
|
|
217
|
|
|
218 #Goes through the features to find which are genes
|
|
|
219
|
|
|
220 for feature in features:
|
|
|
221 gene = ""
|
|
|
222 if feature.type == "gene":
|
|
|
223 locus = "".join(feature.qualifiers["locus_tag"])
|
|
|
224 if "gene" in feature.qualifiers:
|
|
|
225 gene = "".join(feature.qualifiers["gene"])
|
|
|
226 else:
|
|
|
227 continue
|
|
|
228
|
|
|
229 #Goes through the fitness scores of insertions within each gene, and removes whatever % of blank fitness scores were requested along with their corresponding weights
|
|
|
230
|
|
|
231 sum = 0
|
|
|
232 num = 0
|
|
|
233 avgsum = 0
|
|
|
234 blank_ws = 0
|
|
|
235 i = 0
|
|
|
236 if locus in gene_summary.keys():
|
|
|
237 for w in gene_summary[locus]["w"]:
|
|
|
238 if float(w) == 0:
|
|
|
239 blank_ws += 1
|
|
|
240 else:
|
|
|
241 sum += float(w)
|
|
|
242 num += 1
|
|
|
243 count = num + blank_ws
|
|
|
244 removed = 0
|
|
|
245 to_remove = int(float(arguments.blank_pc)*count)
|
|
|
246 if blank_ws > 0:
|
|
|
247 i = 0
|
|
|
248 while i < len(gene_summary[locus]["w"]):
|
|
|
249 if removed == to_remove:
|
|
|
250 break
|
|
|
251 if not w:
|
|
|
252 del gene_summary[locus]["w"][i]
|
|
|
253 del gene_summary[locus]["s"][i]
|
|
|
254 removed += 1
|
|
|
255 i -= 1
|
|
|
256 i += 1
|
|
|
257
|
|
|
258 #If all the fitness values within a gene are empty, sets mean/var to 0.10 and Xs out sd/se; marks the gene if that's requested
|
|
|
259
|
|
|
260 if num == 0:
|
|
|
261 if (arguments.marked and locus in marked_set):
|
|
|
262 output.append([locus, "0.10", "0.10", "X", "X", gene, count, blank_ws, num, removed, "M", "\n"])
|
|
|
263 else:
|
|
|
264 output.append([locus, "0.10", "0.10", "X", "X", gene, count, blank_ws, num, removed, "\n"])
|
|
|
265
|
|
|
266 #Otherwise calls average() or weighted_average() to find the aggregate w / count / standard deviation / standard error of the insertions within each gene; marks the gene if that's requested
|
|
|
267
|
|
|
268 else:
|
|
|
269 if not arguments.weighted:
|
|
|
270 (average, variance, stdev, stderr) = average(gene_summary[locus]["w"])
|
|
|
271 else:
|
|
|
272 (average, variance, stdev, stderr) = weighted_average(gene_summary[locus]["w"],gene_summary[locus]["s"])
|
|
|
273 if (arguments.marked and locus in marked_set):
|
|
|
274 output.append([locus, average, variance, stdev, stderr, gene, count, blank_ws, num, removed, "M", "\n"])
|
|
|
275 else:
|
|
|
276 output.append([locus, average, variance, stdev, stderr, gene, count, blank_ws, num, removed, "\n"])
|
|
|
277
|
|
|
278 #If a gene doesn't have any insertions, sets mean/var to 0.10 and Xs out sd/se, plus leaves count through removed blank because there were no reads.
|
|
|
279
|
|
|
280 else:
|
|
|
281 if (arguments.marked and locus in marked_set):
|
|
|
282 output.append([locus, "0.10", "0.10", "X", "X", gene, "", "", "", "", "M", "\n"])
|
|
|
283 else:
|
|
|
284 output.append([locus, "0.10", "0.10", "X", "X", gene, "", "", "", "", "\n"])
|
|
|
285
|
|
|
286 #Writes the aggregated fitness file
|
|
|
287
|
|
|
288 with open(arguments.summary, "wb") as csvfile:
|
|
|
289 writer = csv.writer(csvfile)
|
|
|
290 writer.writerows(output)
|
|
|
291
|
|
|
292 #If finding missing genes is not requested, just finds the aggregate w / count / standard deviation / standard error of the insertions within each gene, and writes them to a file, plus marks the genes requested
|
|
|
293 #This is never called through Galaxy since finding missing genes is better than not finding them.
|
|
|
294
|
|
|
295 else:
|
|
|
296 output = [["Locus","W","Count","SD","SE","M\n"]]
|
|
|
297 for gene in gene_summary.keys():
|
|
|
298 sum = 0
|
|
|
299 num = 0
|
|
|
300 average = 0
|
|
|
301 if "w" not in gene_summary[gene]:
|
|
|
302 continue
|
|
|
303 for i in gene_summary[gene]["w"]:
|
|
|
304 sum += i
|
|
|
305 num += 1
|
|
|
306 average = sum/num
|
|
|
307 xminusxbars = 0
|
|
|
308 for i in w:
|
|
|
309 xminusxbars += (i-average)**2
|
|
|
310 if num > 1:
|
|
|
311 sd = math.sqrt(xminusxbars/(num-1))
|
|
|
312 se = sd / math.sqrt(num)
|
|
|
313 if (arguments.marked and locus in marked_set):
|
|
|
314 output.append([gene, average, num, sd, se, "M", "\n"])
|
|
|
315 else:
|
|
|
316 output.append([gene, average, num, sd, se, "\n"])
|
|
|
317 with open(arguments.summary, "wb") as csvfile:
|
|
|
318 writer = csv.writer(csvfile)
|
|
|
319 writer.writerows(output)
|
|
|
320
|
|
|
321
|
|
|
322 #Test: python ../script/aggregate.py -m tigr4_normal.txt -w 1 -x 10 -l 50 -b 0 -c NC_003028b2.gbk -o aggregates/L3_2394eVI_GlucTEST.csv results/L3_2394eVI_Gluc.csv
|
|
|
323
|
|
|
324 #Perl Test: perl ../script/aggregate.pl -m tigr4_normal.txt -w 1 -x 10 -l 50 -b 0 -c NC_003028b2.gbk -o aggregates/L3_2394eVI_Gluc.csv results/L3_2394eVI_Gluc.csv |
