def create_SQ_header(database_v,species):
'''
:param database_v: database version
:return: list of chromosomes with their size (from ENSEMBL)
'''
SQ=[]
# create connection to ensembm database
# create connection to ensembl database
if species=="arabidopsis_thaliana":
Genome_species=Species.getCommonName(species.replace('_',' '))
account=host.HostAccount(host="mysql-eg-publicsql.ebi.ac.uk",user="******",passwd="",port=4157)
ensembl=Genome(Species=Genome_species,account=account,Release=30)
else:
Genome_species=Species.getCommonName(species.replace('_',' '))
ensembl=Genome(Species=Genome_species,Release=database_v,account=None)
# convert IDs
coord_table=ensembl.CoreDb.getTable('coord_system')
seq_region_table=ensembl.CoreDb.getTable('seq_region')
select_obj=[seq_region_table.c.name,
seq_region_table.c.length,
coord_table.c.version,
]
from_obj=seq_region_table.join(coord_table,coord_table.c.coord_system_id==seq_region_table.c.coord_system_id)
query = sql.select(select_obj,from_obj=[from_obj],
whereclause = coord_table.c.rank==1)
for row in query.execute():
if '_' not in row[0]:
SQ_string= "@SQ\tSN:chr"+str(row[0])+"\tLN:"+str(row[1])+"\tAS:"+str(row[2])+"\tSP:"+str(species)
SQ.append(SQ_string)
return SQ
def prepareAnnotationENSEMBL(psm_protein_id,mode,database_v,species,three_frame_translation):
'''
:param psm_protein_id: list of protein IDs (untagged)
:param mode: transcript or protein mode
:param database_v: database version
:param species: species name
:return: dictionairy mapping proteins into ENSEMBL
'''
print('Commencing ENSEMBL data retrieval')
# create connection to ensembl database
if species=="arabidopsis_thaliana":
Genome_species=Species.getCommonName(species.replace('_',' '))
account=host.HostAccount(host="mysql-eg-publicsql.ebi.ac.uk",user="******",passwd="",port=4157)
ensembl=Genome(Species=Genome_species,account=account,Release=30)
else:
Genome_species=Species.getCommonName(species.replace('_',' '))
ensembl=Genome(Species=Genome_species,Release=database_v,account=None)
# convert IDs
translation_table=ensembl.CoreDb.getTable('translation')
transcript_table=ensembl.CoreDb.getTable('transcript')
select_obj=[transcript_table.c.stable_id,
translation_table.c.stable_id,
transcript_table.c.transcript_id,
translation_table.c.seq_start,
translation_table.c.start_exon_id,
]
from_obj=translation_table.join(transcript_table,transcript_table.c.transcript_id==translation_table.c.transcript_id)
if mode=='protein':
id=1
query = sql.select(select_obj,from_obj=[from_obj],
whereclause = translation_table.c.stable_id.in_(psm_protein_id))
elif mode=='transcript':
id=0
query = sql.select(select_obj,from_obj=[from_obj],
whereclause = transcript_table.c.stable_id.in_(psm_protein_id))
psm_protein_id={}
transcript_ids=[]
for row in query.execute():
#print row
transcript_ids.append(row[2])
psm_protein_id[row[id]]={'transcript_id':row[0],'translation_id':row[1],
'transcript_seq':'','protein_seq':'',
'chr':'','strand':'','5UTR_offset':row[3],'start_exon_rank':row[4]}
return ensembl_construct_sequences(psm_protein_id,ensembl,transcript_ids,database_v,species,
three_frame_translation,mode)
def setupGenome( species, db_host=None, db_user=None, db_pass=None, db_release=None ):
"""
Notes:
setup the ensembl_genome object using pycogent, this object has methods described here: http://pycogent.org/examples/query_ensembl.html
Args:
species = 'Anopheles gambiae' #string: this needs to match the mysql databases in vbdev: >mysql -hlocalhost -uvbuser -pSavvas
db_host = 'localhost' #string: ^^
db_user = '******' #string: ^^
db_pass = '******' #string: ^^
db_release = 73 #integer: the realease versions we use are all 73
"""
if not db_host:
account=None
if not db_release:
db_release=73
from cogent.db.ensembl import HostAccount, Genome, Species
account = HostAccount(db_host,db_user,db_pass)
species_latin = species.replace('_',' ')
Species.amendSpecies(species_latin,species_latin)
genome = Genome(Species=species_latin,Release=db_release,account=account)
return genome
from __future__ import absolute_import, print_function
import numpy as np
import pandas as pd
from . import base
try:
import cogent
from cogent.db.ensembl import HostAccount, Genome, Compara, Species
account = None
release = '87'
#add recent ensembl species
species = [
'cow', 'human', 'mouse', 'rat', 'chimp', 'gorilla', 'orangutan',
'macaque', 'dog', 'pig', 'cat', 'olive baboon', 'sheep'
]
Species.amendSpecies('Papio anubis', 'olive baboon')
Species.amendSpecies('Ovis aries', 'sheep')
Species.amendSpecies('Erinaceus europaeus', 'hedgehog')
Species.amendSpecies('Mustela putorius furo', 'ferret')
except Exception as e:
print(e)
def get_orthologs(refgenome, ensid=None, symbol=None):
'''if ensid!=None:
mygene = cow.getGeneByStableId(StableId=ensid)
else:
mygene = cow.getGenesMatching(symbol=symbol)'''
#print mygene
orthologs = comp.getRelatedGenes(gene_region=mygene,
Relationship='ortholog_one2one')
def download_database_pycogent(species, release, database_name='ensembl', nucleotide=False) :
log = get_log()
#try :
import cogent
from cogent.db.ensembl import Species, Genome, Compara, HostAccount
from cogent.db.ensembl.database import Database
#except ImportError :
# log.fatal("pycogent import failed, exiting...")
# exit(1)
if cogent.version_info != (1,5,3) :
log.warning("only tested with pycogent version 1.5.3 (you are running %s)" % cogent.version)
release, db_name, db_details = get_missing_info(species, release, database_name)
account = HostAccount(
db_details['hostname'],
db_details['username'],
db_details['password'],
port=db_details['port'])
if Species.getSpeciesName(species) == 'None' : # this is not an error, it returns the string "None"
log.warning("%s not found in pycogent, attempting to add it manually" % species)
Species.amendSpecies(species.capitalize().replace('_', ' '), species)
genome = Genome(species, Release=release, account=account)
compara = Compara([species], Release=release, account=account)
# DON'T TRY THIS AT HOME!
#
# what happens is it searches for compara databases, but unfortunately finds more than one
# in this situation pycogent just connects to the first one, which is always compara_bacteria
# so one solution is to dig through all the compara objects internals to provide a connection
# to the correct database ... obviously not the best solution, but at 6 lines of code definitely
# the shortest ;-P
#
if db_name not in ('ensembl', 'bacteria') :
log.warning("accessing compara from pycogent with species outside of ensembl-main and ensembl-bacteria is problematic, attempting to patch...")
from cogent.db.ensembl.host import DbConnection
from cogent.db.ensembl.name import EnsemblDbName
import sqlalchemy
new_db_name = EnsemblDbName(compara.ComparaDb.db_name.Name.replace('bacteria', db_name))
compara.ComparaDb._db = DbConnection(account=account, db_name=new_db_name)
compara.ComparaDb._meta = sqlalchemy.MetaData(compara.ComparaDb._db)
# end of DON'T TRY THIS AT HOME!
genes = set()
families = []
stderr.write("\r[downloading %s] got %d sequences " % ("CDS" if nucleotide else "protein", len(genes)))
for gene in genome.getGenesMatching(BioType='protein_coding') :
stableid = gene.StableId
# ignore genes that have already been seen as members of other gene families
if stableid in genes :
continue
genes.add(stableid)
paralogs = compara.getRelatedGenes(StableId=stableid, Relationship='within_species_paralog')
current = []
if paralogs is None :
stderr.write("\r[downloading %s] got %d sequences " % ("CDS" if nucleotide else "protein", len(genes)))
current.append((stableid, str(gene.CanonicalTranscript.Cds) if nucleotide else str(gene.CanonicalTranscript.ProteinSeq)))
else :
for paralog in paralogs.Members :
paralogid = paralog.StableId
genes.add(paralogid)
stderr.write("\r[downloading %s] got %d sequences " % ("CDS" if nucleotide else "protein", len(genes)))
try :
current.append((paralogid, str(paralog.CanonicalTranscript.Cds) if nucleotide else str(paralog.CanonicalTranscript.ProteinSeq)))
except AttributeError :
log.fatal("pycogent did not find a canonical transcript for %s" % paralogid)
exit(1)
#print ','.join([ i for i,j in current ])
families.append(current)
stderr.write("\r[downloading %s] got %d sequences\n" % ("CDS" if nucleotide else "protein", len(genes)))
return families
