/
make_plots.py
160 lines (135 loc) · 6.46 KB
/
make_plots.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
"""
Calculates plots
"""
import os
import sys
sys.path.extend(['/Users/ifiddes/hive/ihategit/pipeline/', '/Users/ifiddes/hive/ihategit/pipeline/submodules', '/Users/ifiddes/hive/ihategit/pipeline/submodules/pycbio', '/Users/ifiddes/hive/ihategit/pipeline/submodules/comparativeAnnotator'])
import argparse
import matplotlib
matplotlib.use('Agg')
matplotlib.rcParams['pdf.fonttype'] = 42
import pandas as pd
import seaborn as sns
import matplotlib.pyplot as plt
from math import log
from collections import Counter, defaultdict
from pycbio.sys.fileOps import iterRows, ensureDir, iterLines, ensureFileDir
from pycbio.sys.procOps import runProc, callProcLines
def parse_args():
parser = argparse.ArgumentParser()
parser.add_argument('--attrs', default="/Users/ifiddes/hive/ihategit/pipeline/gencode_vm8/C57B6J.tsv")
parser.add_argument('--counts_dir', default='counts_tables_deseq')
parser.add_argument('--deseq_dir', default='deseq_results')
parser.add_argument('--genomes', nargs='+', default=None)
parser.add_argument('--ref_dir', default='star_alignments_v4/mm10')
parser.add_argument('--tgt_dir', default='star_alignments_v4/strain_specific')
parser.add_argument('--out_dir', default='scatterplots')
parser.add_argument('--experiment', required=True)
return parser.parse_args()
def get_tx_map(attrs):
tx_map = {}
for x in iterRows(attrs, skipLines=1):
tx_map[x[3]] = x[0]
return tx_map
def get_common_name_map(attrs):
common_name_map = {}
for x in iterRows(attrs, skipLines=1):
common_name_map[x[1]] = x[0]
return common_name_map
def load_dir(data, mode, path, genomes, common_name_map):
for g in genomes:
for base_dir, d, files in os.walk(os.path.join(path, g)):
if files:
files = [os.path.join(base_dir, x) for x in files if x.endswith('.filtered.bam.counts.cds')]
for p in files:
for l in iterLines(p, skipLines=2):
gene_name, chrom, start, end, strand, length, count = l.split()
if mode == 'ref':
gene_name = common_name_map[gene_name]
data[g][mode][gene_name].append(count)
def filter_data(d):
ref_txs = {ens_id for ens_id, val in d['ref'].iteritems()}
tgt_txs = {ens_id for ens_id, val in d['tgt'].iteritems()}
keep_names = ref_txs & tgt_txs
r = {}
for n in keep_names:
r[n] = [d['ref'][n], d['tgt'][n]]
return r
def construct_counts_tables(data, base_path):
for g, d in data.iteritems():
counts_path = os.path.join(base_path, g + '.counts.txt')
conds_path = os.path.join(base_path, g + '.conds.txt')
ensureFileDir(counts_path)
cleaned_d = filter_data(d)
with open(counts_path, 'w') as counts:
first_line = cleaned_d.values()[0]
assert len(first_line[0]) == len(first_line[1])
header = ['ref{}'.format(i) for i in range(len(first_line[0]))] + ['tgt{}'.format(i) for i in range(len(first_line[0]))]
counts.write('\t'.join(header) + "\n")
for ens_id, vals in cleaned_d.iteritems():
f = [item for sublist in vals for item in sublist]
l = [ens_id] + f
counts.write('\t'.join(l) + '\n')
with open(conds_path, 'w') as outf:
outf.write('condition\n')
for x in header:
outf.write('\t'.join([x, ''.join([q for q in x if q.isalpha()])]) + "\n")
def strip_version(df):
df.index = [x.split('.')[0] for x in df.index]
return df
def load_paralogy(g, df, para_dir='paralogy_filtered_sets'):
unique_genes = [x.rstrip() for x in open(os.path.join(para_dir, g + '.txt'))]
gene_r = {gene_id: 'Not Unique' for gene_id in df.index}
for gene_id in unique_genes:
gene_r[gene_id] = 'Unique'
return df.join(pd.Series(gene_r, name='Paralogy'), how='inner')
def calculate_pval_vector(df, ref_col_name, tgt_col_name, col='Unnamed: 2'):
num_sig = len([x for x in df[col] if x <= 0.05])
num_non_sig = len(df) - num_sig
df['ref'] = map(log, df['ref'])
df['tgt'] = map(log, df['tgt'])
p_val_vector = ['<= 0.05 (n = {})'.format(num_sig) if x <= 0.05 else '> 0.05 (n = {})'.format(num_non_sig) for x in df[col]]
df[col] = p_val_vector
df.columns = [ref_col_name, tgt_col_name, 'Adjusted p-value']
return df
def generate_plots(genomes, deseq_path, out_dir):
for g in genomes:
base_out = os.path.join(out_dir, g)
ensureDir(base_out)
main_plot = os.path.join(base_out, 'nonzero_log.pdf')
para_plot = os.path.join(base_out, 'nonzero_log_paralogy.pdf')
ref_col_name = 'Normalized counts (aligned to mm10)'
tgt_col_name = 'Normalized counts (aligned to {})'.format(g)
df = pd.read_table(os.path.join(deseq_path, g + '.data_table.tsv'), sep=' ', header=0, index_col=0)
df = df[df['ref'] >= 1]
df = df[df['tgt'] >= 1]
df = strip_version(df)
df = calculate_pval_vector(df, ref_col_name, tgt_col_name)
p = sns.lmplot(x=ref_col_name, y=tgt_col_name, data=df, scatter_kws={"s": 6, "alpha": 0.7}, hue='Adjusted p-value', fit_reg=False)
p.set(xlim=(0, 10), ylim=(0, 10))
p.savefig(main_plot, format='pdf')
df2 = load_paralogy(g, df)
p2 = sns.lmplot(x=ref_col_name, y=tgt_col_name, data=df2, scatter_kws={"s": 6, "alpha": 0.7}, col='Paralogy', hue='Adjusted p-value', fit_reg=False)
p2.set(xlim=(0, 10), ylim=(0, 10))
p2.savefig(para_plot, format='pdf')
plt.close('all')
def main():
args = parse_args()
data = defaultdict(lambda: defaultdict(lambda: defaultdict(list)))
ref_path = os.path.join(args.ref_dir, args.experiment)
tgt_path = os.path.join(args.tgt_dir, args.experiment)
if args.genomes is None:
args.genomes = os.listdir(ref_path)
common_name_map = get_common_name_map(args.attrs)
load_dir(data, 'ref', ref_path, args.genomes, common_name_map)
load_dir(data, 'tgt', tgt_path, args.genomes, common_name_map)
base_counts_dir = os.path.join(args.counts_dir, args.experiment)
construct_counts_tables(data, base_counts_dir)
deseq_path = os.path.join(args.deseq_dir, args.experiment)
ensureDir(deseq_path)
for g in args.genomes:
runProc(['Rscript', 'run_DEseq.R', g, deseq_path, base_counts_dir])
out_dir = os.path.join(args.out_dir, args.experiment)
generate_plots(args.genomes, deseq_path, out_dir)
if __name__ == '__main__':
main()