def parse_input_file(db_output_path, pheno_input_path, gencode_input_path, pb8k_callback):
logging.info("Opening PB8K pheno file")
file_iterator = Utilities.CSVFileIterator(pheno_input_path, delimiter="\t", header=TF1.HEADER, compressed=True)
file_iterator.iterate(pb8k_callback)
logging.info("Opening gencode file")
def fixed_row(gencode, row):
F = Utilities.WDBIF
return (row[F.SNP], gencode.ensemble_version, row[F.GENE_NAME], row[F.REFERENCE_ALLELE], row[F.EFFECT_ALLELE], row[F.WEIGHT], row[F.N_SNP], row[F.GENE_R2], row[F.GENE_PVALUE], row[F.GENE_QVALUE])
class FixGeneCallback(object):
def __init__(self, genes):
self.genes = genes
self.selected = {}
def __call__(self, gencode):
if gencode.name in self.genes:
rows = self.genes[gencode.name]
self.selected[gencode.name] = [fixed_row(gencode, row) for k,row in rows.iteritems()]
gencode_callback = FixGeneCallback(pb8k_callback.genes)
GencodeFile.parse_gencode_file(gencode_input_path, gencode_callback)
genes = gencode_callback.selected
logging.info("Fixing rows")
genes = fix_row(genes)
logging.info("Saving entries")
connection = Utilities.connect(db_output_path)
Utilities.setup_db(connection)
Utilities.insert_entries(connection, genes)
def run(self):
if os.path.exists(self.args.output_file):
logging.info("DB already there, delete it if you want it done again")
return
connection = Utilities.connect(self.args.output_file)
Utilities.setup_db(connection)
parse_input_file(PB8KFileInfo.TF1, connection, self.args.input_file, self.args.gencode_file, self.args.fdr_filter,
self.args.use_variance, self.args.sample_size, self.args.only_best_snp)
def run(self):
if os.path.exists(self.args.output):
logging.info("%s already exists, delete it if you want ity done again", self.args.output)
return
db = Utilities.connect(self.args.output)
Utilities.setup_db(db)
parse_folder(self.args.input_folder, db, self.args.gencode_file)
def process(args):
logging.info("Loading LCTF %s", args.lctf_path)
lctf_entries = LCFileInfo.load_lctf(args.lctf_path)
logging.info("Filling snp rsid from %s", args.gtex_snp_path)
lctf_entries = LCFileInfo.fill_snp_id_from_gtex(lctf_entries, args.gtex_snp_path, expect_throw=args.expect_throw)
logging.info("Building db entries")
#This will drop the previous stuff.
weight_method= WEIGHT_METHODS[args.weight_type]
name_method = NAME_METHODS[args.functional_unit_type]
lctf_entries = LCFileInfo.entries_to_weight_db(lctf_entries, name_method, weight_method)
logging.info("Building db")
connection = Utilities.connect(args.output_file)
Utilities.setup_db(connection)
logging.info("Inserting entries")
Utilities.insert_entries(connection, lctf_entries)
logging.info("Ran successfully")
def parse_input_file(db_output_path, gtex_pheno_input_path, gencode_input_path, gtex_snp_path, gtex_callback):
logging.info("Opening GTEx pheno file %s", os.path.basename(os.path.normpath(gtex_pheno_input_path)))
file_iterator = Utilities.CSVFileIterator(gtex_pheno_input_path, delimiter="\t", header=GTEXEQTLF.HEADER, compressed=True)
file_iterator.iterate(gtex_callback)
logging.info("%d found at GTEx file", len(gtex_callback.genes))
logging.info("Opening gencode file")
def gencode_fixed_row(gencode, row):
F = Utilities.WDBIF
return (row[F.SNP], row[F.GENE], gencode.gene_name, row[F.REFERENCE_ALLELE], row[F.EFFECT_ALLELE], row[F.WEIGHT], row[F.N_SNP], row[F.GENE_R2], row[F.GENE_PVALUE], row[F.GENE_QVALUE])
class FixGeneCallback(object):
def __init__(self, genes):
self.genes = genes
self.selected = {}
def __call__(self, gencode):
if gencode.gene_id in self.genes:
rows = self.genes[gencode.gene_id]
self.selected[gencode.gene_id] = {k:gencode_fixed_row(gencode, row) for k,row in rows.iteritems()}
gencode_callback = FixGeneCallback(gtex_callback.genes)
GencodeFile.parse_gencode_file(gencode_input_path, gencode_callback)
genes = gencode_callback.selected
logging.info("%d survived after gencode file", len(genes))
pvalues = gtex_callback.pvalues
del gencode_callback
del gtex_callback
logging.info("Opening GTEX Snp file")
def snp_fixed_row(row, rsid, ref_allele, eff_allele):
F = Utilities.WDBIF
return (rsid, row[F.GENE], row[F.GENE_NAME], ref_allele, eff_allele, row[F.WEIGHT], row[F.N_SNP], row[F.GENE_R2], row[F.GENE_PVALUE], row[F.GENE_QVALUE])
class FixSNPCallback(object):
def __init__(self, genes):
self.genes = genes
self.selected = {}
self.variant_to_gene = {}
for gene, rows in self.genes.iteritems():
for variant, row in rows.iteritems():
self.variant_to_gene[variant] = gene
def __call__(self, i, comps):
if i == 0:
return
F = GTExSNPFile.SNPTF
variant = comps[F.VariantID]
ref = comps[F.Ref_b37]
eff = comps[F.Alt]
snp = comps[F.RS_ID_dbSNP142_CHG37p13]
if variant in self.variant_to_gene:
gene = self.variant_to_gene[variant]
row = self.genes[gene][variant]
if not gene in self.selected:
self.selected[gene] = {}
self.selected[gene][snp] = snp_fixed_row(row, snp, ref, eff)
snp_callback = FixSNPCallback(genes)
snp_iterator = Utilities.CSVFileIterator(gtex_snp_path, delimiter="\t", compressed=True) #header in the file is just wrong., header=GTExSNPFile.SNPTF.HEADER, compressed=True)
snp_iterator.iterate(snp_callback)
genes = snp_callback.selected
logging.info("%d survived after snp file", len(genes))
del snp_callback
logging.info("Fixing rows")
fix_rows(genes, pvalues)
logging.info("Saving entries")
connection = Utilities.connect(db_output_path)
Utilities.setup_db(connection)
Utilities.insert_entries(connection, genes)
