-
Notifications
You must be signed in to change notification settings - Fork 0
/
add_condition_to_pheno.py
236 lines (205 loc) · 8.47 KB
/
add_condition_to_pheno.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
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
#!/usr/bin/env python3
"""
This script will
- open pheno file
- open genotype file
- open conditional files
- create new pheno files with genotype as new column
these files will be used as new inputs for the
SAIGE-QTL conditional analysis pipeline.
To run:
In main:
analysis-runner \
--description "add variant to pheno files" \
--dataset "bioheart" \
--access-level "standard" \
--output-dir "saige-qtl/bioheart_n787_and_tob_n960/241008_ashg/input_files/" \
python3 add_condition_to_pheno.py \
--pheno-files-path=gs://cpg-bioheart-main/saige-qtl/bioheart_n787_and_tob_n960/241008_ashg/input_files/pheno_cov_files/ \
--conditional-files-path=gs://cpg-bioheart-main-analysis/saige-qtl/bioheart_n787_and_tob_n960/241008_ashg/input_files/conditional_files/ \
--chrom-mt-files-path=gs://cpg-bioheart-main/saige-qtl/bioheart_n787_and_tob_n960/241008_ashg/input_files/genotypes/vds-tenk10k1-0_qc_pass --max-delay=30
In test:
analysis-runner \
--description "add variant to pheno files" \
--dataset "bioheart" \
--access-level "test" \
--output-dir "saige-qtl/bioheart_n990_and_tob_n1055/241004_n100/input_files/" \
python3 add_condition_to_pheno.py \
--pheno-files-path gs://cpg-bioheart-test/saige-qtl/bioheart_n990_and_tob_n1055/241004_n100/input_files/pheno_cov_files/ \
--condition-pheno-files-path gs://cpg-bioheart-test/saige-qtl/bioheart_n990_and_tob_n1055/241004_n100/input_files/pheno_cov_files_condition/ \
--conditional-files-path gs://cpg-bioheart-test-analysis/saige-qtl/bioheart_n990_and_tob_n1055/241004_n100/output_files/conditioning_on_top_one_variant_from_cv_test_round1/ \
--chrom-mt-files-path gs://cpg-bioheart-test/saige-qtl/bioheart_n990_and_tob_n1055/241004_n100/input_files/genotypes/vds-tenk10k1-0_subset
"""
import click
import logging
import pandas as pd
import hail as hl
import hailtop.batch.job as hb_job
from cpg_utils import to_path
from cpg_utils.hail_batch import get_batch, init_batch
def add_variant_to_pheno_file(
mt_path: str,
gene: str,
chrom: str,
celltype: str,
original_pheno_file: str,
pheno_new_file,
conditional_files_path: str,
):
"""
Add variant as column to pheno cov file
"""
import pandas as pd
from cpg_utils.hail_batch import init_batch
init_batch()
# open original pheno cov file
print(f'original pheno file: {original_pheno_file}')
pheno_df = pd.read_csv(original_pheno_file, sep='\t')
# open conditional file to extract variant(s)
conditional_file = f'{conditional_files_path}{celltype}_conditional_file.tsv'
condition_df = pd.read_csv(conditional_file, sep='\t')
conditional_string = condition_df[condition_df['gene'] == gene][
'variants_to_condition_on'
]
variant = conditional_string.values[0]
# extract variant(s) from chrom mt
chrom_mt_filename = f'{mt_path}/{chrom}_common_variants.mt'
chrom_mt = hl.read_matrix_table(chrom_mt_filename)
# extract genotypes for relevant variant(s)
chrom_mt_filtered = chrom_mt.filter_rows(
(chrom_mt.locus.position == int(variant.split(':')[1]))
& (chrom_mt.alleles[1] == variant.split(':')[3])
)
# steal all the entries as a Table, dropping everything except chr, pos, alleles, Genotypes
genos = chrom_mt_filtered.select_entries('GT').select_rows().entries()
# create an integer representation of the genotypes
genos = genos.annotate(
GT=hl.case()
.when(genos.GT.is_hom_var(), 2)
.when(genos.GT.is_het(), 1)
.default(0)
)
# export
genos.export(str(pheno_new_file).replace('.tsv', '_tmp.tsv'), delimiter='\t')
geno_df = pd.read_csv(str(pheno_new_file).replace('.tsv', '_tmp.tsv'), sep='\t')
# rename useful columns and drop the rest
geno_df = geno_df.rename(
columns={'s': 'individual', 'GT': f"chr{variant.replace(':', '_')}"}
)
geno_df = geno_df.drop(['locus', 'alleles'], axis=1)
# add as column to new df (merging on pheno file)
new_pheno_df = pd.merge(geno_df, pheno_df, on='individual')
with pheno_new_file.open('w') as npf:
new_pheno_df.to_csv(npf, index=False, header=True, sep='\t')
@click.command()
@click.option('--pheno-files-path')
@click.option('--condition-pheno-files-path')
@click.option('--conditional-files-path')
@click.option('--chrom-mt-files-path')
@click.option('--ngenes-to-test', default='all')
@click.option(
'--concurrent-job-cap',
type=int,
default=100,
help=(
'To avoid resource starvation, set this concurrency to limit '
'horizontal scale. Higher numbers have a better walltime, but '
'risk jobs that are stuck (which are expensive)'
),
)
@click.option(
'--gene-condition-storage',
default='8G',
)
@click.option(
'--gene-condition-memory',
default='8G',
)
def main(
pheno_files_path: str,
condition_pheno_files_path: str,
conditional_files_path: str,
chrom_mt_files_path: str,
ngenes_to_test: str,
concurrent_job_cap: int,
gene_condition_storage: str,
gene_condition_memory: str,
):
"""
Make conditional pheno cov file for conditional analysis
"""
init_batch()
all_jobs: list[hb_job.Job] = []
def manage_concurrency(new_job: hb_job.Job):
"""
Manage concurrency, so that there is a cap on simultaneous jobs
Args:
new_job (hb_job.Job): a new job to add to the stack
"""
if len(all_jobs) > concurrent_job_cap:
new_job.depends_on(all_jobs[-concurrent_job_cap])
all_jobs.append(new_job)
# extract conditional files using glob
files = [
str(file)
for file in to_path(conditional_files_path).glob('*_conditional_file.tsv')
]
# determine what cell type we have conditional files for
celltypes = [
file.replace(conditional_files_path, '').replace('_conditional_file.tsv', '')
for file in files
]
# loop over celltypes
for celltype in celltypes:
print(f'celltype: {celltype}')
# pheno cov file path
pheno_files_path_ct = f'{pheno_files_path}{celltype}/'
# open cell type specific conditional file
conditional_files_path_ct_file = (
f'{conditional_files_path}{celltype}_conditional_file.tsv'
)
conditional_df = pd.read_csv(conditional_files_path_ct_file, sep='\t')
# extract unique chromosomes
conditional_df['chr'] = [
f"chr{variant.split(':')[0]}" for variant in conditional_df['top_MarkerID']
]
# loop over chromosomes
for chrom in conditional_df['chr'].unique():
print(f'chrom: {chrom}')
conditional_df_chr = conditional_df[conditional_df['chr'] == chrom]
genes = conditional_df_chr['gene']
logging.info(f'genes to test: {", ".join(genes)}')
# if specified, only test ngenes genes
if ngenes_to_test != 'all':
genes = genes[0 : int(ngenes_to_test)]
logging.info(f'I found these files: {", ".join(files)}')
for gene in genes:
print(f'gene: {gene}')
pheno_original_file = (
f'{pheno_files_path_ct}{chrom}/{gene}_{celltype}_pheno_cov.tsv'
)
pheno_new_file = f'{condition_pheno_files_path}{celltype}/{chrom}/{gene}_{celltype}_pheno_cov.tsv'
if not to_path(pheno_new_file).exists():
pheno_cond_job = get_batch().new_python_job(
name=f'gene make pheno cond file: {celltype},{gene}'
)
pheno_cond_job.storage(gene_condition_storage)
pheno_cond_job.memory(gene_condition_memory)
pheno_cond_job.call(
add_variant_to_pheno_file,
mt_path=chrom_mt_files_path,
gene=gene,
chrom=chrom,
celltype=celltype,
original_pheno_file=pheno_original_file,
conditional_files_path=conditional_files_path,
pheno_new_file=to_path(pheno_new_file),
)
manage_concurrency(pheno_cond_job)
logging.info(
f'make conditional pheno cov file job for {gene}, {celltype} scheduled'
)
get_batch().run(wait=False)
if __name__ == '__main__':
logging.basicConfig(level=logging.INFO)
main() # pylint: disable=no-value-for-parameter