def _get_seq_motif(self, refseq_id, residue, pos_str):
seq = self.seq_dict[refseq_id]
pos_1ix = int(pos_str)
pos_0ix = pos_1ix - 1
if seq[pos_0ix] != residue:
self.invalid_site_pos.append((refseq_id, residue, pos_str))
if seq[pos_0ix + 1] == residue:
self.off_by_one.append((refseq_id, residue, pos_str))
motif, respos = \
ProtMapper.motif_from_position_seq(seq, pos_1ix + 1,
self.motif_window)
return {
'site_motif': {
'motif': motif,
'respos': respos,
'off_by_one': True
}
}
else:
return {}
else:
# The index of the residue at the start of the window
motif, respos = ProtMapper.motif_from_position_seq(
seq, pos_1ix, self.motif_window)
return {
'site_motif': {
'motif': motif,
'respos': respos,
'off_by_one': False
}
}
def count_kin_sub_mapped(human_peptides, site_dict):
mapped_sites = []
in_human = 0
not_in_human = 0
annot = 0
total_annot = 0
import random
random.shuffle(human_peptides)
for ix, (hum_up_id, peptide, site_pos) in enumerate(human_peptides):
print(ix)
pm = ProtMapper()
ms = pm.map_peptide_to_human_ref(hum_up_id, 'uniprot', peptide,
site_pos)
if ms.valid:
in_human += 1
# Check if the human site is annotated
site_key = (ms.up_id, ms.mapped_res, ms.mapped_pos)
if site_key in site_dict:
annot += 1
num_kinases = len(site_dict[site_key])
total_annot += num_kinases
else:
not_in_human += 1
#print("Not in human:", hum_up_id, peptide, site_pos)
mapped_sites.append(ms)
print("Not hum", not_in_human, "hum", in_human, "annot", annot,
"total_annot", total_annot)
return mapped_sites
def map_bel_sites():
with open(BEL_AGENTS, 'rb') as f:
bel_agents = pickle.load(f)
pm = ProtMapper(use_cache=True, cache_path=CACHE_PATH)
bel_sites = map_agents(BEL_AGENTS, pm, 'bel')
with open(BEL_SITES, 'wb') as f:
pickle.dump(bel_sites, f)
pm.save_cache()
def build_data(data_sites, dm_opt):
pm = ProtMapper()
filt_sites = []
for site in data_sites:
(refseq, gene, res, pos, pep, respos) = site
hgnc_name, up_id = up_for_hgnc(gene)
if dm_opt is True and up_id is not None:
ms = pm.map_peptide_to_human_ref(up_id, 'uniprot', pep,
int(respos) + 1)
if ms.valid and ms.mapped_res and ms.mapped_pos:
res = ms.mapped_res
pos = ms.mapped_pos
site = (refseq, hgnc_name, res, pos, pep, respos)
filt_sites.append(site)
return filt_sites
def valid_counts(sitelist):
results = []
pm = ProtMapper()
for ix, (refseq, gene, res, pos, pep, respos) in enumerate(sitelist):
result = {'refseq': refseq, 'gene': gene, 'res': res, 'pos': pos,
'peptide': pep, 'respos': respos}
if ix % 10000 == 0:
print(ix)
up_rs = up_id_for_rs(refseq)
hgnc_name, up_hgnc = up_for_hgnc(gene)
result['up_hgnc'] = up_hgnc
result['up_rs'] = up_rs
if up_rs is None:
result['up_rs_iso_specific'] = None
else:
result['up_rs_iso_specific'] = iso_specific(up_rs)
hgnc_map_result = map_peptide(up_hgnc, res, pos, pep, respos, 'hgnc',
pm)
rs_map_result = map_peptide(up_rs, res, pos, pep, respos, 'rs',
pm)
result.update(hgnc_map_result)
result.update(rs_map_result)
results.append(result)
df = pd.DataFrame.from_dict(results, orient='columns')
return df
def map_sites(sites_dict):
"""Tabulate valid, invalid, and mapped sites from a set of Agents."""
site_map = {}
pm = ProtMapper()
for site_ix, site in enumerate(sites_dict.keys()):
if site_ix % 1000 == 0:
print('%d of %d' % (site_ix, len(sites_dict)))
up_id, res, pos = site
try:
ms = pm.map_to_human_ref(up_id, 'uniprot', res, pos)
site_map[site] = ms
except Exception as e:
logger.exception(e)
logger.info("up_id: %s, res %s, pos %s" % (up_id, res, pos))
# Now that we've collected a list of all the sites, tabulate frequencies
return site_map
def map_pc_sites():
pm = ProtMapper(use_cache=True, cache_path=CACHE_PATH)
agent_files = {
'hprd': 'output/biopax/PathwayCommons10.hprd.BIOPAX.pkl',
'kegg': 'output/biopax/PathwayCommons10.kegg.BIOPAX.pkl',
'panther': 'output/biopax/PathwayCommons10.panther.BIOPAX.pkl',
'pid': 'output/biopax/PathwayCommons10.pid.BIOPAX.pkl',
'psp_pc': 'output/biopax/PathwayCommons10.psp.BIOPAX.pkl',
'pc_tsv': 'output/psp_kinase_substrate_tsv.pkl',
'reactome': 'output/biopax/PathwayCommons10.reactome.BIOPAX.pkl',
'wp': 'output/biopax/PathwayCommons10.wp.BIOPAX.pkl',
'psp_biopax': 'output/biopax/Kinase_substrates.pkl',
#'reactome_human': 'output/biopax/Homo_sapiens.pkl',
}
all_sites = {}
for db_name, agent_file in agent_files.items():
sites = map_agents(agent_file, pm, db_name)
all_sites[db_name] = sites
with open(BIOPAX_SITES_BY_DB, 'wb') as f:
pickle.dump(all_sites, f)
pm.save_cache()
def _get_seq_motif(self, refseq_id, residue, pos_str):
seq = self.seq_dict[refseq_id]
pos_1ix = int(pos_str)
pos_0ix = pos_1ix - 1
if seq[pos_0ix] != residue:
self.invalid_site_pos.append((refseq_id, residue, pos_str))
if seq[pos_0ix + 1] == residue:
self.off_by_one.append((refseq_id, residue, pos_str))
motif, respos = \
ProtMapper.motif_from_position_seq(seq, pos_1ix + 1,
self.motif_window)
return {'site_motif': {'motif': motif, 'respos': respos,
'off_by_one': True}}
else:
return {}
else:
# The index of the residue at the start of the window
motif, respos = ProtMapper.motif_from_position_seq(seq, pos_1ix,
self.motif_window)
return {'site_motif': {'motif': motif, 'respos': respos,
'off_by_one': False}}
import json
from flask import Flask, request, abort, Response, jsonify
from flask_cors import CORS
from protmapper import ProtMapper
app = Flask(__name__)
CORS(app)
optional_bool_args = {
'do_methionine_offset': True,
'do_orthology_mapping': True,
'do_isoform_mapping': True
}
pm = ProtMapper()
@app.route('/map_to_human_ref', methods=['GET', 'POST'])
def map_to_human_ref():
required_args = ('prot_id', 'prot_ns', 'residue', 'position')
# Require all required arguments
for arg in required_args:
if request.json.get(arg) is None:
abort(Response('The required argument "%s" is missing.' % arg,
400))
# Now set the required arguments and the optional ones with default values
# as backup
arg_values = {key: request.json.get(key) for key in required_args}
for arg, default_value in optional_bool_args.items():
value = request.json.get(arg, default_value)
arg_values[arg] = value