def test1(self):
self.setup()
# To retrieve IDs and relvancy scores from pubmed for PMID 9298984
# to the pubmed database:
# dbfrom=pubmed&id=9298984&cmd=neighbor
results = self.client.from_dbids(EUtils.DBIds(
"pubmed", "9298984")).neighbor_links()
def test2(self):
self.setup()
# To retrieve IDs from nucleotide for GI 18250303, 18250301, 18250299
# to protein:
# dbfrom=nucleotide&db=protein&id=18250303,18250307
records = self.client.from_dbids(
EUtils.DBIds("nucleotide", ["18250303", "18250301", "18250299"]))
neighbors = records.neighbor_links("protein")
def testRelatedSequences(self):
# Get all protein sequences similar to 4579714 (bacteriorhodopsin)
# in FASTA format
client = HistoryClient.HistoryClient(eutils=picklestore.client())
result = client.post(EUtils.DBIds("protein", "4579714"))
related = result.neighbor_links("protein")
related_dbids = related.linksetdbs["protein_protein"].dbids
proteins = client.post(related_dbids)
fasta_infile = proteins.efetch(retmode="text", rettype="fasta")
fasta = fasta_infile.read()
print fasta[:1000]
def test3(self):
self.setup()
# To retrieve pubmed related articles for PMIDs 11812492 11774222
# with a publication date from 1995 to the present:
# dbfrom=pubmed&id=11812492,11774222&db=pubmed&mindate=1995&datetype=pdat
# NOTE: this example is wrong -- it need the current date too!
records = self.client.from_dbids(
EUtils.DBIds("pubmed", ["11812492", "11774222"]))
# Can't actually use this since that would cause the regression
# tests to fail.
today = time.strftime("%Y/%m/%d")
today = "2003/01/12"
related = records.neighbor_links(
daterange=EUtils.DateRange("1995", today, "pdat"))
def testRelatedItems(self):
# Get proteins similar to 4579714 (bacteriorhodopsin) which
# were published in 2002
client = DBIdsClient.DBIdsClient(eutils=picklestore.client())
results = client.from_dbids(EUtils.DBIds("protein", "4579714"))
neighbors = results.neighbor_links("protein",
daterange=EUtils.DateRange(
"2002/01/01", "2002/12/31",
"pdat"))
dbids = neighbors.linksetdbs["protein_protein"].dbids
print
print len(
dbids), "sequences similar to GI:4579714 were published in 2002"
print "The record identifiers are:", ", ".join(map(str, dbids))
print client.from_dbids(dbids).efetch(retmode="text",
rettype="summary").read()
def testSeqFetch(self):
entrez = HistoryClient.HistoryClient(eutils=picklestore.client())
dbids = EUtils.DBIds("nucleotide", "8886082")
result = entrez.post(dbids)
lines = result.efetch(retmode="text", rettype="gb").readlines()
_trim_lines(lines)
self.assertEquals(lines[-2], """\
2761 ttcctagata aacac
""")
lines = result.efetch(retmode="text",
rettype="gb",
seq_start=1,
seq_stop=60).readlines()
_trim_lines(lines)
self.assertEquals(
lines[-2], """\
1 ttcagtttat ttggacggaa gaatggtggc tcaattattg acgacttgtg gcctaaattt
""")
lines = result.efetch(retmode="text",
rettype="gb",
seq_start=3,
seq_stop=40,
strand=EUtils.PLUS_STRAND).readlines()
_trim_lines(lines)
self.assertEquals(
lines[-2], """\
1 cagtttattt ggacggaaga atggtggctc aattattg
""")
lines = result.efetch(retmode="text",
rettype="gb",
seq_start=3,
seq_stop=40,
strand=EUtils.MINUS_STRAND).readlines()
_trim_lines(lines)
self.assertEquals(
lines[-2], """\
1 caataattga gccaccattc ttccgtccaa ataaactg
""")
def testForNewItems(self):
# See if there are Halobacterium salinarum hits
known_dbids = EUtils.DBIds(
'protein',
[
# '461608', '133739', '121859', '121858', '114808',
'26399547',
'21362599',
'20807968',
'18310162',
'10954592',
'7674074',
'3913880',
'3913879',
'3913877',
'1350915',
'462366',
'462365',
'461784',
'119171',
'21617825',
'18202987',
'14423949',
'13878670',
'13431821',
'13124532',
'12230579',
'7531150',
'6094165',
'1352354',
'1350833',
'1169376',
'140149',
'133039',
'133038',
'132873',
'132841',
'114547',
'114516',
'18138454',
'11385314',
'11385312',
'11385311',
'11383975',
'11383843',
'11383705',
'11383685',
'11383631',
'11383600',
'11383544',
'11383521',
'11383449',
'11383413',
'11362758',
'11362757',
'11362756',
'11362633',
'11361208',
'11361204',
'11361203',
'11361202',
'11360968',
'11360864',
'11360863',
'11360531',
'11291672',
'11290031',
'11283208',
'11281612',
'11280741',
'11280740',
'11280737',
'11280563',
'11280281',
'11280223',
'11279456',
'11277771',
'11277752',
'11277727',
'11277613',
'11277612',
'11277611',
'11277609',
'11277608',
'11277607',
'11277606',
'11277605',
'11277604',
'11277603',
'11277602',
'11277601',
'11277597',
'11277160',
'11277109',
'11277100',
'11271301',
'11270360',
'11261072',
'11260624',
'11255573',
'9972746',
'7443396',
'7440307',
'7427602',
'2129406',
'2129405',
'2129404',
'2117914',
'1363464',
'1363463',
'1084274',
'1076149',
'1076148',
'1076147',
'629399',
'629398',
'629396',
'629394',
'487075',
'487074',
'486684',
'477959',
'477726',
'477027',
'421715',
'421714',
'282664',
'282663',
'282662',
'282661',
'282660',
'282659',
'282658',
'282657',
'282656',
'282655',
'282654',
'282653',
'282652',
'281164',
'281162',
'281161',
'281160',
'281159',
'281158',
'281157',
'280348',
'99214',
'99212',
'99211',
'99210',
'99209',
'99207',
'99206',
'99205',
'99203',
'99202',
'99201',
'99200',
'99198',
'99195',
'99192',
'99191',
'99190',
'99188',
'99187',
'99185',
'99183',
'81076',
'81071',
'81070',
'81069',
'81067',
'81065',
'81063',
'81062',
'81061',
'81060',
'81058',
'81057',
'81055',
'81054',
'81052',
'81051',
'81047',
'81045',
'81043',
'81042',
'81041',
'81040',
'81039',
'81038',
'81037',
'81036',
'81035',
'81034',
'81032',
'81031',
'81029',
'81026',
'81023',
'81022',
'81020',
'81019',
'81018',
'81017',
'81016',
'81015',
'81013',
'81012',
'81011',
'76372',
'76329',
'76324',
'72632',
'71284',
'71211',
'71179',
'71146',
'71063',
'66382',
'21264507',
'18000395',
'7531067',
'132751',
'117568',
'14916722',
'133992',
'3122880',
'114811',
'24158915',
'24158914',
'24158913',
'10640268',
'3913878',
'3023997',
'544287',
'19909691',
'19909603',
'3183098',
'141254',
'2492920',
'809698',
'43513',
'18144841',
'4467437',
'4467436',
'6226495',
'3122803',
'3122663',
'2499389',
'2494605',
'1350834',
'548760',
'141356',
'141353',
'140804',
'139955',
'134663',
'134662',
'133462',
'133078',
'132785',
'132646',
'120245',
'120232',
'114812',
'20873485',
'20873480',
'17942995',
'17942994',
'17942993',
'20516540',
'20151159',
'20150922',
'20150921',
'12644298',
'12644014',
'120249',
'10120917',
'13124533',
'133398',
'11132078',
'6225334',
'1168858',
'549772',
'140827',
'140825',
'120239',
'114838',
'16974947',
'16974946',
'3219755',
'347255',
'8569313',
'4929918',
'899266',
'10121032',
'1621047',
'14278674',
'14278519',
'11071687',
'11071686',
'11071685',
'11071684',
'6729723',
'6729722',
'6435624',
'6435623',
'3745775',
'12655906',
'12655904',
'12655903',
'6435626',
'6435625',
'11992133',
'8918496',
'10121037',
'10121036',
'10121033',
'10120892',
'10120851',
'10120850',
'2851428',
'1710568',
'1710498',
'548772',
'548769',
'133031',
'132995',
'132985',
'132879',
'132784',
'132750',
'132728',
'132708',
'132639',
'120251',
'7327959',
'7107278',
'6435593',
'455306',
'455305',
'455304',
'455303',
'455302',
'455301',
'455300',
'455299',
'455298',
'455297',
'150418',
'150417',
'6435594',
'6172231',
'6172230',
'6172229',
'6172228',
'6172227',
'6137453',
'130153',
'118549',
'2499384',
'2499383',
'461612',
'461611',
'461610',
'5822280',
'285806',
'4930169',
'2425186',
'2425185',
'2425184',
'2425183',
'2425182',
'2425181',
'2425180',
'2425179',
'2425178',
'2425177',
'2425176',
'2425175',
'2425174',
'3659953',
'3659944',
'4469246',
'4001706',
'4001704',
'4388967',
'4378986',
'4377598',
'809702',
'809701',
'43668',
'43667',
'43666',
'43665',
'43663',
'43662',
'43661',
'43660',
'43658',
'43491',
'4322492',
'4104487',
'4104485',
'4104483',
'3928158',
'43508',
'1633466',
'2072795',
'1527138',
'1527137',
'1353676',
'3015619',
'598123',
'515085',
'493889',
'493888',
'229726',
'2351849',
'2351848',
'1154790',
'1154789',
'1154788',
'1154787',
'1154786',
'1154785',
'1154784',
'1154783',
'1154782',
'1154781',
'1154780',
'1154779',
'1154778',
'1154777',
'2190417',
'2190416',
'984742',
'984741',
'984740',
'984739',
'2760612',
'2648028',
'1070346',
'1070345',
'1070344',
'2209068',
'1487875',
'1199752',
'1199750',
'509675',
'509674',
'285817',
'285810',
'285808',
'216709',
'43455',
'43454',
'43453',
'43452',
'43451',
'43450',
'807110',
'1654427',
'1654425',
'1654423',
'1654421',
'1654419',
'225948',
'1583108',
'1583107',
'1094422',
'448273',
'226716',
'226715',
'350080',
'226309',
'225904',
'225428',
'1483625',
'1235894',
'1435134',
'1435132',
'1435131',
'1435129',
'994803',
'994802',
'226310',
'223370',
'223077',
'223076',
'223063',
'1333716',
'671101',
'671100',
'550341',
'311841',
'297410',
'49046',
'43656',
'43655',
'43654',
'43653',
'43641',
'43640',
'43559',
'43557',
'43552',
'43551',
'43550',
'43548',
'43546',
'43545',
'43544',
'43543',
'43542',
'43541',
'43540',
'43539',
'43537',
'43535',
'43534',
'43533',
'43531',
'43530',
'43526',
'43525',
'43524',
'43523',
'43522',
'43521',
'43520',
'43519',
'43518',
'43517',
'43511',
'43510',
'43505',
'43504',
'43503',
'43501',
'43499',
'43498',
'43496',
'43495',
'43493',
'517390',
'305353',
'305352',
'148816',
'148814',
'148812',
'148794',
'148793',
'148792',
'148768',
'148767',
'148766',
'148764',
'148763',
'148759',
'148757',
'148753',
'148749',
'148747',
'148745',
'305350'
])
# Upload to the server
client = HistoryClient.HistoryClient(eutils=picklestore.client())
old_records = client.post(known_dbids)
# Now see if there's anything new
new_records = client.search("Halobacterium salinarum BUTNOT #%s" %
(old_records.query_key, ),
db="protein")
print
print "There are", len(
new_records), "new Halobacterium salinarum records"
new_dbids = new_records.dbids
print "The record identifiers are:", ", ".join(map(str, new_dbids))
print new_records.efetch(retmode="text", rettype="summary").read()
# These should exist, since I commented them out above :)
for x in ('461608', '133739', '121859', '121858', '114808'):
assert x in new_dbids, "Cannot find expected record %r" % x
def post(ids, entrez_db="protein"):
"Make a request to NCBI, yield each response separately"
result = client().post(EUtils.DBIds("protein", ids))
tree = ET.parse(result.efetch())
for seq_node in tree.findall('.//GBSeq'):
yield seq_node
def setup(self):
self.entrez = HistoryClient.HistoryClient(eutils=picklestore.client(1))
self.result1 = self.entrez.post(EUtils.DBIds("protein", "4579714"))
self.result2 = self.entrez.post(
EUtils.DBIds("nucleotide", ["18250303", "18250301", "18250299"]))
def testClient(self):
eutils = picklestore.client()
infile = eutils.esearch("Dalke",
field="au",
daterange=EUtils.DateRange("1995", "1998"),
retstart=1,
retmax=5,
usehistory=1,
webenv=None)
s = infile.read()
try:
counts = map(int, re.findall(r"<Count>(-?\d+)</Count>", s))
assert len(counts) == 4, counts
assert counts[0] == 10
assert counts[1] >= 30
assert counts[2] == -1
assert counts[3] == -1
assert s.find("<RetMax>5</RetMax>") != -1
assert s.find("<RetStart>1</RetStart>") != -1
ids = re.findall(r"<Id>(\d+)</Id>", s)
assert len(ids) == 5, ids
terms = re.findall(r"<Term>([^<]+)</Term>", s)
assert len(terms) == 3
assert terms[0] == "Dalke[Author]"
assert terms[1] == "1995[EDAT]"
assert terms[2] == "1998[EDAT]"
query_key1 = re.findall(r"<QueryKey>(\d+)</QueryKey>", s)[0]
assert query_key1 == "1", query_key1 # always true?
quoted_webenv = re.findall("<WebEnv>([^>]+)</WebEnv>", s)[0]
webenv = urllib.unquote(quoted_webenv)
except:
print "ERROR!"
print s
raise
try:
# Can I refetch those same Ids using the history?
t = ""
t = eutils.efetch_using_history(db="pubmed",
webenv=webenv,
query_key=query_key1,
retstart=1,
retmax=5,
retmode="text",
rettype="uilist").read()
new_ids = t.split()
assert ids == new_ids, (ids, new_ids) # Must be in same order too!
except:
print "ERROR!"
print s
print " -- and --"
print t
raise
# Make sure I'm getting the same XML summary through history and id
sum1 = sum2 = None
try:
sum1 = eutils.esummary_using_history(db="pubmed",
webenv=webenv,
query_key=query_key1,
retstart=1,
retmax=1).read()
sum2 = eutils.esummary_using_dbids(
dbids=EUtils.DBIds("pubmed", [ids[0]])).read()
assert sum1 == sum2
except:
print "Summary 1"
print sum1
print "-----------------"
print "Summary 2"
print sum2
raise
# Make sure I'm getting the same XML version of the records
rec1 = rec2 = None
try:
rec1 = eutils.efetch_using_history(db="pubmed",
webenv=webenv,
query_key=query_key1,
retmode="xml",
retstart=1,
retmax=1).read()
rec2 = eutils.efetch_using_dbids(dbids=EUtils.DBIds(
"pubmed", [ids[0]]),
retmode="xml").read()
assert rec1 == rec2
except:
print "Record 1"
print rec1
print "-----------------"
print "Record 2"
print rec2
raise
# Post a few GIs (from the protein database) to the server
# This appends to the existing history so should be query_key #2.
post_ids = ["914034", "5263173", "1769808", "1060883"]
infile = eutils.epost(EUtils.DBIds("protein", post_ids), webenv=webenv)
post_results = infile.read()
try:
query_key2 = re.findall(r"<QueryKey>(\d+)</QueryKey>",
post_results)[0]
assert query_key2 == "2"
quoted_webenv = re.findall("<WebEnv>([^>]+)</WebEnv>",
post_results)[0]
webenv = urllib.unquote(quoted_webenv)
except:
print "ERROR"
print post_results
raise
# Verify that the posted ids are correct
posted_ids = eutils.efetch_using_history(
db="pubmed",
webenv=webenv,
query_key=query_key2,
retstart=0,
retmax=len(post_ids),
retmode="text",
rettype="uilist").read().split()
x1 = posted_ids[:] # Make copies since I need the correct
x1.sort() # order for getting the FASTA version, below
x2 = post_ids[:]
x2.sort()
assert x1 == x2, (post_ids, posted_ids)
# Now fetch them as FASTA format
fasta1 = fasta2 = None
try:
fasta1 = eutils.efetch_using_history(db="protein",
webenv=webenv,
query_key=query_key2,
retstart=0,
retmax=len(post_ids),
retmode="text",
rettype="fasta").read()
fasta2 = eutils.efetch_using_dbids(dbids=EUtils.DBIds(
"protein", posted_ids),
retmode="text",
rettype="fasta").read()
assert fasta1 == fasta2
except:
print "ERROR FASTA1"
print fasta1
print "ERROR FASTA2"
print fasta2
raise
# It's much harder to test the ELink capabilities.
# Get the VMD paper
results = None
try:
results = eutils.esearch(
"Humphrey W. AND Dalke A. AND Schulten K. AND VMD[Title]",
field="au").read()
# There should only be one match
ids = re.findall(r"<Id>(\d+)</Id>", results)
assert ids == ["8744570"]
except:
print "Error"
print results
raise
# Look at the related publications and we should find
# my Tcl paper, which is 9390282
links = None
try:
links = eutils.elink_using_dbids(EUtils.DBIds("pubmed", ids),
cmd="neighbor").read()
# remember, the first id comes from the <Id> in <IdList>
related_ids = re.findall(r"<Id>(\d+)</Id>", links)[1:]
assert "9390282" in related_ids
except:
print "Error"
print links
raise
# Get the taxonomy record for the "posted_ids".
# NOTE: This test original compared the 2nd element from
# that list, but elink_using_history doesn't support the
# retstart/retmax parameters.
#
# This comes from query_key2 in the history. Do it both ways
# to compare results.
link1 = link2 = None
try:
link1 = eutils.elink_using_dbids(EUtils.DBIds(
"protein", posted_ids),
db="taxonomy",
cmd="neighbor").read()
link2 = eutils.elink_using_history(dbfrom="protein",
webenv=webenv,
query_key=query_key2,
db="taxonomy",
cmd="neighbor").read()
assert link1 == link2
taxids = re.findall(r"<Id>(\d+)</Id>", link1)[len(posted_ids):]
assert taxids == ["43776", "29282", "28442", "2237"], taxids
except:
print "Error",
print link1
print "----------------"
print link2
raise
# See if there are linkouts 914034
# Should be at least one, to DART 3240.
llinks = None
try:
llinks = eutils.elink_using_dbids(EUtils.DBIds(
"protein", [posted_ids[1]]),
cmd="llinks").read()
assert llinks.find("<ObjUrl>") != -1
assert "3240" in re.findall("<Id>(\d+)</Id>", llinks)
except:
print "ERROR"
print llinks
raise
# Finally, check that I can limit the seach to an Entrez query string
# I'm using the example
# "retrieve MEDLINE indexed only related articles for PMID 12242737"
# elink.fcgi?dbfrom=pubmed&id=12242737&db=pubmed&term=medline[sb]
full = restricted = None
try:
full = eutils.elink_using_dbids(
EUtils.DBIds("pubmed", ["12242737"])).read()
restricted = eutils.elink_using_dbids(EUtils.DBIds(
"pubmed", ["12242737"]),
term="medline[sb]").read()
counts1 = full.count("<Link>")
counts2 = restricted.count("<Link>")
assert counts1 > counts2
except:
print "ERROR"
print full
print "---------"
print restricted
raise