def _combine_beds(beds, combine_mode, merge=False):
if combine_mode == "all":
return BedOperations.intersect(beds, merge=merge).result
elif combine_mode == "any":
return BedOperations.union(beds, merge=merge).result
else:
raise InvalidCombineOperationException
def to_remove_get_tss_fantom5_bed1(genes, tissues, genome, combine_mode, upstream, downstream, *args):
BedsProcessor.logger().info("Extracting F5 TSS for genes (%s): %s; tissues: %s; and combine_mode: %s" % (genome, genes, tissues, combine_mode))
flanked_genes = BedsProcessor._get_gene_promoter_sites(genes, genome, upstream, downstream)
beds = BedsProcessor._get_regulatory_regions_bed(genome,
tissues,
RegulatoryRegionsFilesRegistry.FANTOM5_PROMOTERS_KEY,
flanked_genes.sort().merge())
BedsProcessor.log_count("Flanked genes' promoters BED", flanked_genes)
BedsProcessor.log_bed(flanked_genes)
BedsProcessor.logger().info("F5 TSS BEDs list:\n%s" % str(beds))
if beds and flanked_genes:
joined_f5_tss = BedsProcessor._combine_beds(beds, combine_mode)
BedsProcessor.log_count("Combined F5 TSS BED", joined_f5_tss)
BedsProcessor.log_bed(joined_f5_tss)
result = BedOperations.intersect([joined_f5_tss, flanked_genes], **{"u": True}).result
BedsProcessor.log_count("F5 TSS BED intersected with genes' promoters",result)
BedsProcessor.log_bed(result)
return [result]
else:
BedsProcessor.logger().info("Returning empty TSS list")
return []
def _generic_get_promoters_bed(source, genes, tissues, genome, combine_mode, upstream, downstream, *args):
BedsProcessor.logger().info(f"Extracting {source} for genes ({genome}): {genes}; tissues: {tissues}; and combine_mode: {combine_mode}")
flanked_genes = BedsProcessor._get_gene_promoter_sites(genes, genome, upstream, downstream)
beds = BedsProcessor._get_regulatory_regions_bed(genome, tissues, source, flanked_genes.sort().merge())
BedsProcessor.log_count("Flanked genes' promoters BED", flanked_genes)
BedsProcessor.log_bed(flanked_genes)
BedsProcessor.logger().info(f"{source} BEDs list:\n{beds}")
if beds and flanked_genes:
joined_promoters = BedsProcessor._combine_beds(beds, combine_mode)
BedsProcessor.log_count(f"Combined {source} BED", joined_promoters)
BedsProcessor.log_bed(joined_promoters)
result = BedOperations.intersect([joined_promoters, flanked_genes], **{"u": True}).result
BedsProcessor.log_count(f"{source} BED intersected with genes' promoters", result)
BedsProcessor.log_bed(result)
return [result]
else:
BedsProcessor.logger().info("Returning empty promoters list")
return []
def to_remove_get_enhancers_encode_bed(genes, tissues, genome, combine_mode, upstream, downstream, *args):
BedsProcessor.logger().info("Extracting ENCODE enhancers for genes (%s): %s; tissues: %s; and combine_mode: %s" % (genome, genes, tissues, combine_mode))
flanked_genes = BedsProcessor._get_gene_promoter_sites(genes, genome,
int(float(upstream) * 1000),
int(float(downstream) * 1000))
beds = BedsProcessor._get_regulatory_regions_bed(genome,
tissues,
RegulatoryRegionsFilesRegistry.ENCODE_ENHANCERS_KEY,
flanked_genes.sort().merge())
BedsProcessor.log_count("Flanked genes' promoters BED", flanked_genes)
BedsProcessor.log_bed(flanked_genes)
BedsProcessor.logger().info("ENCODE enhancer BEDs list:\n%s" % str(beds))
if beds and flanked_genes:
joined_enc_enh = BedsProcessor._combine_beds(beds, combine_mode)
BedsProcessor.log_count("Combined ENCODE enhancers BED", joined_enc_enh)
BedsProcessor.log_bed(joined_enc_enh)
result = BedOperations.intersect([joined_enc_enh, flanked_genes], **{"u": True}).result
BedsProcessor.log_count("ENCODE enhancers BED intersected with genes' promoters", result)
BedsProcessor.log_bed(result)
return [ result ]
else:
BedsProcessor.logger().info("Returning empty enhancers list")
return []
def get_bed_fragment(self, tissue, source_symbol, regions):
"""
Get slice of a BED. Filtering on non-tabixed BED files is not supported.
If regions is None, entire BED is returned
"""
self.logger.info('Requested {}tissue [{}] from source [{}]'.format(
"fragment [%s] from " % regions if regions else "", tissue,
source_symbol))
if tissue not in self._available_tissues:
raise InvalidTissueNameException(
"Querried tissue [%s] was not among available tissue keys:\n%s"
% (tissue, str(self._available_tissues.keys())))
try:
bed_path = self._available_tissues[tissue][source_symbol]
track_name = source_symbol + "(" + tissue.split(
'(')[0].strip().replace(" ", "_") + ")"
self.logger.info('Found %s. Adding name %s' %
(bed_path, track_name))
full_bed = BedLoader(bed_path)
if regions is None:
return BedOperations.add_name(full_bed.bed, track_name)
else:
beds = [full_bed.filter_by(i) for i in regions]
if any(beds):
filtered_bed = BedOperations.union([e for e in beds if e],
merge=False).result
return BedOperations.add_name(filtered_bed, track_name)
except KeyError:
self.logger.info('No tissue [%s] in source [%s]' %
(tissue, source_symbol))
return None
def get_genes_bed(genes, genome, *args):
BedsProcessor.logger().info("Querying gene database for %s" % genes)
registry = GenesDBRegistry.get_instance()
gene_beds = [ registry.get_bed(genome, genes) ]
BedsProcessor.log_count("Result BED file", gene_beds)
BedsProcessor.log_bed(gene_beds)
result = BedOperations.union(gene_beds).result
BedsProcessor.log_count("Union of the BED files", result)
BedsProcessor.log_bed(result)
return [ result ]
def _get_accessible_mirnas(mirna_symbols, tissues, genome, combine_mode):
mirna_bed = BedsProcessor.get_genes_bed(mirna_symbols, genome)
# mirna_bed is one element list
if mirna_bed[0] is None:
return None
# intersect beds with accessible chromatin in tissues
accessible_chromatin = BedsProcessor._get_regulatory_regions_bed(genome, tissues,
RegulatoryRegionsFilesRegistry.ENCODE_CHROMATIN_KEY,
mirna_bed[0].sort().merge())
if any(accessible_chromatin):
accessible_chromatin_aggregate = BedsProcessor._combine_beds(accessible_chromatin, combine_mode)
accessible_mirna = BedOperations.intersect(mirna_bed + [accessible_chromatin_aggregate], merge=False).result
return accessible_mirna
return None
def perform():
try:
start_time = time.time()
params = get_perform_params()
collected_beds_map = BedsCollector(params).collect_bed_files()
collected_beds = [
bed for beds_list in collected_beds_map.values()
for bed in beds_list
]
final_result = BedOperations.union(collected_beds, merge=True, **{"c": 4, "o": "distinct"}).result \
if len(collected_beds) > 1 else collected_beds[0]
tmp_file_path = save_as_tmp(final_result)
session["last_result"] = tmp_file_path.name
end_time = (time.time() - start_time)
return return_summary(final_result, end_time)
except Exception as e:
app.logger.exception("Error occurred, details:")
return "Error occurred"
def _get_gene_promoter_sites(genes, genome, upstream, downstream):
genes_bed = BedsProcessor.get_genes_bed(genes, genome)[0]
promoters = BedOperations.get_promoter_region(genes_bed, upstream, downstream)
return promoters.result