-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcal_AI_4pop.py
159 lines (137 loc) · 7.7 KB
/
cal_AI_4pop.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
# -*- coding: utf-8 -*-
"""
Created on Thu Nov 26 15:35:14 2020
@author: YudongCai
@Email: [email protected]
"""
import allel
import click
import numpy as np
import pandas as pd
def select_samples(samples_all, samples_query):
"""
从samples_all里面筛选samples_query，产生布尔掩码
"""
return [True if sample in samples_query else False for sample in samples_all]
def load_vcf2array(vcffile, samples_queried, outsamples: None):
callset = allel.read_vcf(vcffile, samples=samples_queried,
fields=['samples', 'calldata/GT', 'variants/CHROM', 'variants/POS'])
gt_array = callset['calldata/GT'] # 三维array
samples_all = callset['samples']
pos_array = callset['variants/POS']
chrom_array = callset['variants/CHROM']
# 样本是否都在vcf里
if len(samples_queried) != len(samples_all):
samples_notexist = set(samples_queried) - set(samples_all)
print(f'{len(samples_notexist)} samples not exist in the vcf file:')
print(', '.join(samples_notexist))
# 只保留双等位, 多等位以后再考虑
selection_biallelic = np.max(np.max(gt_array, axis=2), axis=1) < 2
gt_array = gt_array[selection_biallelic, :, :]
pos_array = pos_array[selection_biallelic]
chrom_array = chrom_array[selection_biallelic]
n_sites, n_samples, n_hap = gt_array.shape
print(f'{n_sites} biallelic sites were remained.')
# 把outsamples中最高频率的allele设置为alt
if outsamples:
selection_outsamples = select_samples(samples_all, outsamples)
print(f'{np.sum(selection_outsamples)} outgroup samples in vcf file.')
gt_array_out = gt_array[:, selection_outsamples, :].reshape(n_sites, np.sum(selection_outsamples)*n_hap)
selection_swtich = np.sum(gt_array_out==0, axis=1) > np.sum(gt_array_out>0, axis=1) # ref(0)的数量比非ref(!=0)但不是miss(-1)的数量多
print(f'Swtich REF and ALT in {np.sum(selection_swtich)} sites.')
assert gt_array.min() >= -1
assert gt_array.max() <= 1
gt_swtich = gt_array[selection_swtich, :, :]
gt_swtich[gt_swtich == 1] = 9
gt_swtich[gt_swtich == 0] = 1
gt_swtich[gt_swtich == 9] = 0
gt_array[selection_swtich, :, :] = gt_swtich
return gt_array, callset['samples'], pos_array, chrom_array
def cal_alt1_freq(gt_array):
"""
gt_array为load_vcf2array产生的3维ndarray
注意，返回的是第一个ALT的frequency
"""
return allel.GenotypeArray(gt_array).count_alleles().to_frequencies()[:, 1]
@click.command()
@click.option('--vcffile', help='输入的vcf文件')
@click.option('--popa', help='群体A(非渗入群体)的ID列表，一行一个')
@click.option('--popb', help='群体B(被渗入群体)的ID列表，一行一个')
@click.option('--popc', help='群体C(候选的渗入来源群体)的ID列表，一行一个')
@click.option('--popd', help='群体D(需要排除的渗入来源群体)的ID列表，一行一个')
@click.option('--binwidth', type=int, default=50000, help='滑动窗口大小')
@click.option('--stepsize', type=int, default=10000, help='滑动窗口步长')
@click.option('--outprefix', help='输出文件前缀')
def main(vcffile, popa, popb, popc, popd, binwidth, stepsize, outprefix):
"""
U_A,B,C(w,x,y)
A是非渗入群体，B是被渗入群体，C是渗入来源群体
在窗口内A中频率小于w，B中大于x，C中大于y的SNP位点数即为U_A,B,C(w,x,y)
详见：Signatures of Archaic Adaptive Introgression in Present-Day Human Populations
"""
samples_popA = [x.strip() for x in open(popa)]
samples_popB = [x.strip() for x in open(popb)]
samples_popC = [x.strip() for x in open(popc)]
samples_popD = [x.strip() for x in open(popd)]
samples_queried = samples_popA + samples_popB + samples_popC + samples_popD
assert len(set(samples_queried)) == len(samples_queried), "样本ID有重复"
gt_array, samples_all, pos_array, chrom_array = load_vcf2array(vcffile, samples_queried, outsamples=samples_popC)
# 群体C和D中的频率
selection_popC = select_samples(samples_all, samples_popC)
af_popC = cal_alt1_freq(gt_array[:, selection_popC, :]) # 这个返回的是第一个ALT的frequency
selection_popD = select_samples(samples_all, samples_popD)
af_popD = cal_alt1_freq(gt_array[:, selection_popD, :]) # 这个返回的是第一个ALT的frequency
# 只保留群体C中频率百分百且在D中百分之零的位点, ALT按popC转换过了，所以直接算ALT的频率
selection_freq_popC = af_popC == 1 # 在popC中频率100%
selection_freq_popD = af_popD == 0 # 在popD中频率为0%
selection_freq_popCD = selection_freq_popC & selection_freq_popD
gt_array = gt_array[selection_freq_popCD, :, :]
pos_array = pos_array[selection_freq_popCD]
chrom_array = chrom_array[selection_freq_popCD]
af_popC = af_popC[selection_freq_popCD]
af_popD = af_popD[selection_freq_popCD]
# 群体AB的频率，使用筛选后的位点计算
selection_popA = select_samples(samples_all, samples_popA)
af_popA = cal_alt1_freq(gt_array[:, selection_popA, :]) # 这个返回的是第一个ALT的frequency
selection_popB = select_samples(samples_all, samples_popB)
af_popB = cal_alt1_freq(gt_array[:, selection_popB, :]) # 这个返回的是第一个ALT的frequency
# 先把频率文件保存一下
df = pd.DataFrame({'chrom': chrom_array, 'pos': pos_array, 'popA': af_popA, 'popB': af_popB, 'popC': af_popC, 'popD': af_popD})
df.to_csv(f'{outprefix}_altFreq.tsv.gz', index=False, compression='gzip', sep='\t', float_format='%.3f')
print('freq file saved.')
# 过滤满足指标要求的位点
selection_popA_1percent = df['popA'].values < 0.01
df = df.iloc[selection_popA_1percent, :]
selection_popC_1percent = df['popC'].values == 1
df = df.iloc[selection_popC_1percent, :]
selection_popD_1percent = df['popD'].values == 0
df = df.iloc[selection_popD_1percent, :]
print(f'{df.shape[0]} sites remained after frequency filtering.')
# 计算相关统计量
selection_popB_10percent = df['popB'].values >= 0.1
selection_popB_20percent = df['popB'].values >= 0.2
selection_popB_50percent = df['popB'].values >= 0.5
selection_popB_80percent = df['popB'].values >= 0.8
# 滑动窗口计算统计量
odf = []
for offset in range(0, binwidth, stepsize):
df[f'bin_index'] = ((df['pos'].values - 1) - offset) // binwidth
for group_name, gdf in df.groupby(by=['chrom', 'bin_index']):
chrom, bin_index = group_name
start = bin_index * binwidth + offset + 1
if start < 0: # 开头几个窗口长度不足的就直接跳过
continue
end = start + binwidth - 1
n_snp = gdf.shape[0]
print(chrom, start, end, n_snp)
# Q_1_100_q90, Q_1_100_q95, Q_1_100_q100 = np.quantile(gdf['popB'].values, [0.9, 0.95, 1]) # A和B群体的频率已经提前过滤了 New in version 1.15.0.
Q_1_100_q90, Q_1_100_q95, Q_1_100_q100 = np.percentile(gdf['popB'].values, [90, 95, 100])
U_1_10_100 = np.sum(gdf['popB'].values >= 0.1)
U_1_20_100 = np.sum(gdf['popB'].values >= 0.2)
U_1_50_100 = np.sum(gdf['popB'].values >= 0.5)
U_1_80_100 = np.sum(gdf['popB'].values >= 0.8)
odf.append([chrom, start, end, n_snp, Q_1_100_q90, Q_1_100_q95, Q_1_100_q100, U_1_10_100, U_1_20_100, U_1_50_100, U_1_80_100])
odf = pd.DataFrame(odf, columns=['chrom', 'start', 'end', 'n_snp', 'Q90', 'Q95', 'Q100', 'U10', 'U20', 'U50', 'U80']).sort_values(by=['chrom', 'start'])
odf.to_csv(f'{outprefix}_stat.tsv.gz', index=False, compression='gzip', sep='\t', float_format='%.3f')
if __name__ == '__main__':
main()