def test_get_evaluation_tasks_on_dir_with_lower_stages_dirs(self):
temp_dir = tempfile.mkdtemp()
try:
params = D3RParameters()
dataimport = DataImportTask(temp_dir, params)
blast = BlastNFilterTask(temp_dir, params)
os.mkdir(os.path.join(temp_dir, dataimport.get_dir_name()))
os.mkdir(os.path.join(temp_dir, blast.get_dir_name()))
stf = EvaluationTaskFactory(temp_dir, params)
task_list = stf.get_evaluation_tasks()
self.assertEquals(len(task_list), 0)
finally:
shutil.rmtree(temp_dir)
def test_run_with_blast_success_postanalysis_success_no_summary_file(self):
temp_dir = tempfile.mkdtemp()
try:
params = D3RParameters()
params.blastnfilter = 'true'
params.postanalysis = '/bin/echo'
params.pdbdb = '/pdbdb'
blasttask = BlastNFilterTask(temp_dir, params)
blasttask._can_run = True
blasttask.run()
self.assertEqual(blasttask.get_status(), D3RTask.COMPLETE_STATUS)
self.assertEqual(blasttask.get_error(), None)
complete_file = os.path.join(blasttask.get_dir(),
D3RTask.COMPLETE_FILE)
self.assertEqual(os.path.isfile(complete_file), True)
std_err_file = os.path.join(blasttask.get_dir(), 'echo.stderr')
self.assertEqual(os.path.isfile(std_err_file), True)
std_out_file = os.path.join(blasttask.get_dir(), 'echo.stdout')
dataimport = DataImportTask(temp_dir, params)
f = open(std_out_file, 'r')
echo_out = f.read().replace('\n', '')
echo_out.index('--compinchi ' +
os.path.join(temp_dir, dataimport.get_dir_name(),
DataImportTask.COMPINCHI_ICH))
echo_out.index(' ' +
os.path.join(temp_dir, blasttask.get_dir_name()))
f.close()
self.assertEqual(os.path.isfile(std_out_file), True)
self.assertEquals(blasttask.get_status(), D3RTask.COMPLETE_STATUS)
self.assertEquals(
os.path.exists(os.path.join(blasttask.get_dir(),
'true.stderr')), True)
self.assertEquals(
os.path.exists(os.path.join(blasttask.get_dir(),
'true.stdout')), True)
res = blasttask.get_email_log().rstrip('\n')
res.index('/bin/echo')
res.index('# txt files found: 0')
res.index('Output from summary.txt')
finally:
shutil.rmtree(temp_dir)
def test_can_run_does_not_exist_or_error(self):
temp_dir = tempfile.mkdtemp()
try:
params = D3RParameters()
task = DataImportTask(temp_dir, params)
# no make blast db
self.assertEquals(task.can_run(), False)
self.assertEquals(task.get_error(),
'makeblastdb task has notfound status')
self.assertEquals(task._can_run, False)
make_blast = MakeBlastDBTask(temp_dir, params)
make_blast.create_dir()
# make blast db failed
err_file = os.path.join(make_blast.get_dir(), D3RTask.ERROR_FILE)
open(err_file, 'a').close()
self.assertEquals(task.can_run(), False)
self.assertEquals(task.get_error(),
'makeblastdb task has error status')
self.assertEquals(task._can_run, False)
os.remove(err_file)
# make blast db success
open(os.path.join(make_blast.get_dir(), D3RTask.COMPLETE_FILE),
'a').close()
self.assertEquals(task.can_run(), True)
self.assertEquals(task.get_error(), None)
self.assertEquals(task._can_run, True)
task.create_dir()
open(os.path.join(task.get_dir(), D3RTask.ERROR_FILE), 'a').close()
self.assertEquals(task.can_run(), False)
self.assertEquals(task._can_run, False)
self.assertEquals(
task.get_error(),
task.get_dir_name() + ' already exists and ' + 'status is ' +
D3RTask.ERROR_STATUS)
finally:
shutil.rmtree(temp_dir)
def test_get_sequence_tsv(self):
params = D3RParameters()
task = DataImportTask('/foo', params)
self.assertEqual(
task.get_sequence_tsv(), '/foo/' + task.get_dir_name() +
'/new_release_structure_sequence_canonical.tsv')
def test_get_nonpolymer_tsv(self):
params = D3RParameters()
task = DataImportTask('/foo', params)
self.assertEqual(
task.get_nonpolymer_tsv(), '/foo/' + task.get_dir_name() +
'/new_release_structure_nonpolymer.tsv')
def test_get_participant_list_csv(self):
params = D3RParameters()
task = DataImportTask('/foo', params)
self.assertEqual(
task.get_participant_list_csv(), '/foo/' + task.get_dir_name() +
'/' + DataImportTask.PARTICIPANT_LIST_CSV)
def test_get_crystalph_tsv(self):
params = D3RParameters()
task = DataImportTask('/foo', params)
self.assertEqual(
task.get_crystalph_tsv(), '/foo/' + task.get_dir_name() +
'/new_release_crystallization_pH.tsv')
def test_run_with_blast_success_useoldseq_and_postanalysis_fail(self):
temp_dir = tempfile.mkdtemp()
try:
params = D3RParameters()
params.blastnfilter = '/bin/echo'
params.postanalysis = os.path.join(temp_dir, 'foo.py')
params.pdbdb = '/pdbdb'
blasttask = BlastNFilterTask(temp_dir, params)
blasttask._can_run = True
txt_file = os.path.join(blasttask.get_dir(), 'summary.txt')
txt_contents = ('INPUT SUMMARY\\n' + ' sequences: 177\\n' +
' complexes: 149\\n')
# create fake blastnfilter script that makes csv files
f = open(params.postanalysis, 'w')
f.write('#! /usr/bin/env python\n\n')
f.write('f = open(\'' + txt_file + '\', \'w\')\n')
f.write('f.write(\'' + txt_contents + '\\n\')\n')
f.write('f.flush()\nf.close()\n')
f.flush()
f.close()
os.chmod(params.postanalysis, stat.S_IRWXU)
blasttask.run()
self.assertEqual(blasttask.get_status(), D3RTask.COMPLETE_STATUS)
self.assertEqual(blasttask.get_error(), None)
complete_file = os.path.join(blasttask.get_dir(),
D3RTask.COMPLETE_FILE)
self.assertEqual(os.path.isfile(complete_file), True)
std_err_file = os.path.join(blasttask.get_dir(), 'echo.stderr')
self.assertEqual(os.path.isfile(std_err_file), True)
std_out_file = os.path.join(blasttask.get_dir(), 'echo.stdout')
dataimport = DataImportTask(temp_dir, params)
makeblast = MakeBlastDBTask(temp_dir, params)
f = open(std_out_file, 'r')
echo_out = f.read().replace('\n', '')
echo_out.index('--nonpolymertsv ' +
os.path.join(temp_dir, dataimport.get_dir_name(),
DataImportTask.NONPOLYMER_TSV))
echo_out.index(' --sequencetsv ' +
os.path.join(temp_dir, dataimport.get_dir_name(),
DataImportTask.OLDSEQUENCE_TSV))
echo_out.index(' --pdbblastdb ' +
os.path.join(temp_dir, makeblast.get_dir_name()))
echo_out.index(' --compinchi ' +
os.path.join(temp_dir, dataimport.get_dir_name(),
DataImportTask.COMPINCHI_ICH))
echo_out.index(' --outdir ' +
os.path.join(temp_dir, blasttask.get_dir_name()))
echo_out.index(' --crystalpH ' +
os.path.join(temp_dir, dataimport.get_dir_name(),
DataImportTask.CRYSTALPH_TSV))
echo_out.index(' --pdbdb /pdbdb ')
f.close()
self.assertEqual(os.path.isfile(std_out_file), True)
self.assertEquals(blasttask.get_status(), D3RTask.COMPLETE_STATUS)
self.assertEquals(
os.path.exists(
os.path.join(blasttask.get_dir(), 'foo.py.stderr')), True)
self.assertEquals(
os.path.exists(
os.path.join(blasttask.get_dir(), 'foo.py.stdout')), True)
res = blasttask.get_email_log().rstrip('\n')
res.index('/bin/echo')
res.index('# txt files found: 0')
res.index('Output from summary.txt')
res.index(' sequences: 177')
res.index(' complexes: 149')
res.index(dataimport.get_sequence_tsv() +
' file not found falling back to ' +
dataimport.get_oldsequence_tsv())
finally:
shutil.rmtree(temp_dir)
def test_can_run(self):
tempDir = tempfile.mkdtemp()
try:
# try where makeblastdb is not complete
params = D3RParameters()
blastTask = BlastNFilterTask(tempDir, params)
self.assertEqual(blastTask.can_run(), False)
# try where makeblastdb failed
blastDb = MakeBlastDBTask(tempDir, params)
blastDb.create_dir()
errorFile = os.path.join(blastDb.get_path(),
blastDb.get_dir_name(),
D3RTask.ERROR_FILE)
open(errorFile, 'a').close()
self.assertEqual(blastTask.can_run(), False)
self.assertEqual(blastTask.get_error(),
'makeblastdb task has error status')
# try where data import is not complete
completeFile = os.path.join(blastDb.get_path(),
blastDb.get_dir_name(),
D3RTask.COMPLETE_FILE)
open(completeFile, 'a').close()
self.assertEqual(blastTask.can_run(), False)
self.assertEqual(blastTask.get_error(),
'dataimport task has ' + 'notfound status')
# try where data import failed
dataImport = DataImportTask(tempDir, params)
dataImport.create_dir()
errorFile = os.path.join(dataImport.get_path(),
dataImport.get_dir_name(),
D3RTask.ERROR_FILE)
open(errorFile, 'a').close()
self.assertEqual(blastTask.can_run(), False)
self.assertEqual(blastTask.get_error(),
'dataimport task has error status')
# try where blast can run
os.remove(errorFile)
completeFile = os.path.join(dataImport.get_dir(),
D3RTask.COMPLETE_FILE)
open(completeFile, 'a').close()
self.assertEqual(blastTask.can_run(), True)
self.assertEqual(blastTask.get_error(), None)
# try where blast exists
blastTask.create_dir()
self.assertEqual(blastTask.can_run(), False)
self.assertEqual(
blastTask.get_error(),
blastTask.get_dir_name() + ' already exists and' +
' status is unknown')
# try where blast is complete
completeFile = os.path.join(blastTask.get_path(),
blastTask.get_dir_name(),
D3RTask.COMPLETE_FILE)
open(completeFile, 'a').close()
self.assertEqual(blastTask.can_run(), False)
self.assertEqual(blastTask.get_error(), None)
finally:
shutil.rmtree(tempDir)
def main():
blasttask = BlastNFilterTask('', p)
dataimport = DataImportTask('', p)
challenge = ChallengeDataTask('', p)
glide = GlideTask('', p)
makedb = MakeBlastDBTask('', p)
prot = ProteinLigPrepTask('', p)
vina = AutoDockVinaTask('', p)
chimeraprep = ChimeraProteinLigPrepTask('', p)
desc = """
Version {version}
Runs the 9 stages (makedb, import, blast, challengedata,
proteinligprep, {chimeraprep}, extsubmission, glide, vina, &
evaluation) of CELPP processing pipeline
(http://www.drugdesigndata.org)
CELPP processing pipeline relies on a set of directories
with specific structure. The pipeline runs a set of stages
Each stage has a numerical value and a name. The numerical
value denotes order and the stage name identifies separate
tasks to run in the stage.
The filesystem structure of the stage is:
stage.<stage number>.<task name>
The stage(s) run are defined via the required --stage flag.
To run multiple stages serially just pass a comma delimited
list to the --stage flag. Example: --stage import,blast
NOTE: When running multiple stages serially the program will
not run subsequent stages if a task in a stage fails.
Also note order matters, ie putting blast,import will
cause celpprunner.py to run blast stage first.
This program drops a pid lockfile
(celpprunner.<stage>.lockpid) in celppdir to prevent duplicate
invocation.
When run, this program will examine the stage and see
if work can be done. If stage is complete or previous
steps have not completed, the program will exit silently.
If previous steps have failed or current stage already
exists in an error or uncomplete state then program will
report the error via email using addresses set in --email
flag. Errors will also be reported via stderr/stdout.
The program will also exit with nonzero exit code.
This program utilizes simple token files to denote stage
completion. If within the stage directory there is a:
'{complete}' file - then stage is done and no other
checking is done.
'error' file - then stage failed.
'start' file - then stage is running.
Notification of stage start and end will be sent to
addresses set via --email flag.
Unless --customweekdir is set, this program will
examine the 'celppdir' (last argument passed on
commandline) to find the latest directory with this path:
<year>/dataset.week.#
The program will find the latest <year> and within
that year the dataset.week.# with highest #. The output
directories created will be put within this directory.
Setting --customweekdir will cause program to use 'celppdir'
path.
Setting the --createweekdir flag will instruct this
program to create a new directory for the current
celpp week/year before running any stage processing.
NOTE: CELPP weeks start on Friday and end on Thursday
and week # follows ISO8601 rules so week numbers
at the end and start of the year are a bit
wonky.
Breakdown of behavior of program is defined by
value passed with --stage flag:
If --stage '{createchallenge}'
This is NOT a stage, but has the same effect as
calling --stage makedb,import,blast,challengedata
The four stages that need to run to generate the challenge
data package.
If --stage 'makedb'
In this stage the file {pdb_seqres} is downloaded from
an ftp site set by --pdbsequrl.
This file is then gunzipped and NCBI makeblastdb
(set by --makeblastdb) is run on it to create a blast
database. The files are stored in {makeblastdb_dirname}
If --stage 'import'
In this stage 4 files are downloaded from urls specified
by --compinchi and --pdbfileurl flags on the commandline
into {dataimport_dirname} directory.
The tsv files are (--pdbfileurl flag sets url to
download these files from):
{nonpolymer_tsv}
{sequence_tsv}
{crystal_tsv}
The Components ich file is (--compinchi flag sets base url to
download this file from):
{compinchi_ich}
This stage will just wait and retry if any of the tsv files
have NOT been updated since the start of the current
celpp week as determined by a HEAD request. To bypass
this delay add --skipimportwait flag. --importsleep denotes
the time to wait before re-examining the update time of the
tsv files and --importretry sets number of times to retry
before giving up.
If --stage 'blast'
Verifies {dataimport_dirname} exists and has '{complete}'
file. Also verifies {makeblastdb_dirname} exists and has
'{complete}' file. If both conditions are met then the
'blast' stage is run which invokes script set by
--blastnfilter flag and output stored in
{blast_dirname}.
Requires --pdbdb to be set to a directory with valid PDB
database files.
Note: --blastnfilter script is killed after time set with
--blastnfiltertimeout flag.
If --stage 'challengedata'
Verifies {blast_dirname} exists and has '{complete}'
file. If complete, this stage runs which invokes program
set in --genchallenge flag to create a challenge dataset
file. The --pdbdb flag must also be set when calling this
stage. If --ftpconfig is set with {challengepath} field then
this stage will also upload the challenge dataset tarfile
to the ftp server with path set by {challengepath}. The
code will also upload a {latest_txt} file containing name
of the tarfile to the same destination overwriting any
{latest_txt} file that already exists.
Example file for --ftpconfig:
{host} some.ftp.com
{user} bob
{passn} mypass
{path} /celpp
{challengepath} /challenge
{submissionpath} /submissions
If --stage '{chimeraprep}'
Verifies {challenge_dirname} exists and has '{complete}'
file. If complete, this stage runs which invokes program
set in --chimeraprep flag to prepare pdb and inchi files
storing output in {chimeraprep_dirname}. --pdbdb flag
must also be set when calling this stage.
If --stage 'proteinligprep'
Verifies {challenge_dirname} exists and has '{complete}'
file. If complete, this stage runs which invokes program
set in --proteinligprep flag to prepare pdb and inchi files
storing output in {proteinligprep_dirname}. --pdbdb flag
must also be set when calling this stage.
If --stage 'extsubmission'
Connects to server specified by --ftpconfig and downloads
external docking submissions from {submissionpath} on remote
server.
Submissions should be named:
celpp_weekXX_YYYY_dockedresults_ZZZZ.tar.gz as documented here:
https://github.com/drugdata/d3r/wiki/Proposed-challenge-docked\
-results-file-structure
For each submission a directory named stage.X.ZZZZ.extsubmission
will be created and uncompressed contents of package will be
stored in that directory. If data does not conform properly
'error' file will be placed in directory denoting failure
If --stage 'glide'
Verifies {proteinligprep_dirname} exists and has a '{complete}'
file within it. If complete, this stage runs which invokes
program set in --glide flag to perform docking via glide
storing output in {glide_dirname}
If --stage 'vina'
Verifies {proteinligprep_dirname} exists and has a '{complete}'
file within it. If complete, this stage runs which invokes
program set in --vina flag to perform docking via AutoDock Vina
storing output in {vina_dirname}
If --stage 'evaluation'
Finds all stage.{dockstage}.<algo> directories with '{complete}'
files in them which do not end in name '{webdata}' and runs
script set via --evaluation parameter storing the result of
the script into stage.{evalstage}.<algo>.evaluation. --pdbdb flag
must also be set when calling this stage.
""".format(makeblastdb_dirname=makedb.get_dir_name(),
dataimport_dirname=dataimport.get_dir_name(),
blast_dirname=blasttask.get_dir_name(),
challenge_dirname=challenge.get_dir_name(),
createchallenge=CREATE_CHALLENGE,
proteinligprep_dirname=prot.get_dir_name(),
glide_dirname=glide.get_dir_name(),
vina_dirname=vina.get_dir_name(),
dockstage=str(glide.get_stage()),
evalstage=str(glide.get_stage() + 1),
complete=blasttask.COMPLETE_FILE,
chimeraprep_dirname=chimeraprep.get_dir_name(),
chimeraprep=CHIMERA_PREP,
compinchi_ich=DataImportTask.COMPINCHI_ICH,
pdb_seqres=MakeBlastDBTask.PDB_SEQRES_TXT_GZ,
nonpolymer_tsv=DataImportTask.NONPOLYMER_TSV,
sequence_tsv=DataImportTask.SEQUENCE_TSV,
crystal_tsv=DataImportTask.CRYSTALPH_TSV,
webdata=EvaluationTaskFactory.WEB_DATA_SUFFIX,
latest_txt=ChallengeDataTask.LATEST_TXT,
host=FtpFileTransfer.HOST,
user=FtpFileTransfer.USER,
passn=FtpFileTransfer.PASS,
path=FtpFileTransfer.PATH,
challengepath=FtpFileTransfer.CHALLENGEPATH,
submissionpath=FtpFileTransfer.SUBMISSIONPATH,
version=d3r.__version__)
theargs = _parse_arguments(desc, sys.argv[1:])
theargs.program = sys.argv[0]
theargs.version = d3r.__version__
util.setup_logging(theargs)
try:
run_stages(theargs)
except Exception:
logger.exception("Error caught exception")
sys.exit(2)
