Merge branch '35-workflow-job-parsing-update'

Heavily updated the yaml file parsing for two primary reasons:
1. Make it more understandable what's actually going on underneath.
2. Allow for optional arguments.

chipathlon/conf.py

@@ -65,6 +65,11 @@ executables = [
     "chip-job-zcat-peak"
 ]
 
+# Java needs to have -Xmx specified...
+java_tools = [
+    "gem"
+]
+
 # Peak_type validation
 peak_types = {
     "spp": ["narrow", "broad"],
@@ -136,33 +141,33 @@ resources = {
 argument_types = {
     "argument": ["string", "numeric"],
     "file": ["file", "rawfile", "stdout", "stderr"],
-    "list": ["list"],
+    "list": ["file_list", "argument_list"],
     "folder": ["folder", "rawfolder"]
 }
 
 # Defines information about arguments
 argument_keys = {
-    "required": ["type", "changeable", "has_value"],
-    "optional": ["required", "default", "file_type", "path"]
+    "required": ["type", "changeable", "has_value", "required"],
+    "optional": ["default", "path", "file_type"]
 }
 
 # workflow_job keys
 job_keys = {
-    "required": ["inputs", "additional_inputs", "outputs", "command", "arguments"] + resources.keys(),
+    "required": ["inputs", "outputs", "command", "arguments"] + resources.keys(),
     "optional": []
 }
 
+job_inout_keys = {
+    "required": ["type"],
+    "optional": ["file_type"]
+}
+
 # param keys
 param_keys = {
     "required": [],
     "optional": ["arguments"] + resources.keys()
 }
 
-file_list_keys = {
-    "required": ["name", "type"],
-    "optional": ["file_type"]
-}
-
 # workflow order
 workflow = ["align", "remove_duplicates", "peak_calling"]
 

chipathlon/db.py

@@ -6,9 +6,11 @@ import traceback
 import os
 import itertools
 import time
+import collections
 import chipathlon.conf
 from pprint import pprint
 import hashlib
+from chipathlon.utils import progress
 
 
 class MongoDB(object):
@@ -35,7 +37,7 @@ class MongoDB(object):
         self.password = password
         self.client = MongoClient(host)
         self.db = self.client.chipseq
-        self.cache = {}
+        self.cache = collections.defaultdict(dict)
         try:
             self.db.authenticate(username, password, mechanism="SCRAM-SHA-1")
         except:
@@ -45,12 +47,12 @@ class MongoDB(object):
         self.gfs = gridfs.GridFS(self.db)
         return
 
-    def add_cache(self, function, key, data):
+    def add_cache(self, accession, file_type, data):
         """
-        :param function: The function name
-        :type function: str
-        :param key: The key to add the cache entry under.
-        :type key: Any hashable
+        :param accession: The accession of the file to store.
+        :type accession: str
+        :param file_type: The type of file to store.
+        :type file_type: str
         :param data: The data to add to the cache.
         :type data: Object
 
@@ -60,23 +62,20 @@ class MongoDB(object):
         peak calling tools.  In these cases we don't want to request info
         from the database multiple times for the same data.
         """
-        if function not in self.cache:
-            self.cache[function] = {}
-        self.cache[function][key] = data
+        self.cache[accession][file_type] = data
         return
 
-    def get_cache(self, function, key):
+    def get_cache(self, accession, file_type):
         """
-        :param function: The function name
-        :type function: str
-        :param key: The key to get from the cache.
-        :type key: Any hashable
+        :param accession: The accession of the file to retrieve.
+        :type accession: str
+        :param file_type: The type of file to retrieve.
+        :type file_type: str
 
         Gets a data item from the internal cache.
         """
-        if function in self.cache:
-            if key in self.cache[function]:
-                return self.cache[function][key]
+        if accession in self.cache:
+            return self.cache[accession].get(file_type)
         return None
 
     def delete_result(self, result, genome):
@@ -331,7 +330,7 @@ class MongoDB(object):
         valid = True
         msg = ""
         data = {}
-        check_cache = self.get_cache("get_sample", accession)
+        check_cache = self.get_cache(accession, file_type)
         if check_cache is not None:
             msg = "Retrieved data from cache."
             data = check_cache
@@ -342,7 +341,7 @@ class MongoDB(object):
             })
             if cursor.count() == 1:
                 data = cursor.next()
-                self.add_cache("get_sample", accession, data)
+                self.add_cache(accession, file_type, data)
             else:
                 valid = False
                 msg = "Found %s files with accession: %s, file_type: %s. Should only be 1." % (
@@ -352,6 +351,42 @@ class MongoDB(object):
                 )
         return (valid, msg, data)
 
+    def clean_gfs(self):
+        """
+        This function finds all files stored in gridfs that are not currently
+        referenced by any result file and removes them.
+        A clean database is a happy database.
+        """
+        cursor = self.db.results.aggregate([
+            {
+                "$group": {
+                    "_id": 1,
+                    "valid_ids": {"$push": "$gridfs_id"}
+                }
+            }
+        ])
+        # Doc contains all our valid ids
+        id_doc = cursor.next()
+        # Find all fs.files documents
+        gfs_cursor = self.db.fs.files.find({
+            "_id": {
+                "$nin": id_doc["valid_ids"]
+            }
+        })
+        # Iterate through file, delete fs.chunks then fs.files
+        total_files = gfs_cursor.count()
+        print "Found %s unused gridfs files.  Preparing to delete...." % (total_files,)
+        for i, fs_file in enumerate(gfs_cursor):
+            progress(i, total_files)
+            self.db.fs.chunks.remove({
+                "files_id": fs_file["_id"]
+            })
+            self.db.fs.files.remove({
+                "_id": fs_file["_id"]
+            })
+        progress(total_files, total_files)
+        return
+
     def get_samples(self, experiment_accession, file_type):
         """
         :param experiment_accession: Accession number of the experiment to grab samples from.

chipathlon/generators/align_generator.py

@@ -12,17 +12,28 @@ class AlignGenerator(ModuleGenerator):
     :param mdb: A MongoDB database class for fetching sample meta data.
     :type mdb: :py:class:`~chipathlon.db.MongoDB`
     :param workflow_module: The actual module being used.
-    :type workflow_module: :py:class:`~chipathlon.workflow_module.WorkflowModule`
+    :type workflow_module: `~chipathlon.workflow_module.WorkflowModule`
     :param workflow_jobs: Dictionary mapping workflow_job name -> workflow_job instance
     :type workflow_jobs: dict
     :param base_path: Base location of the workflow, used to save metadata files.
     :type base_path: str
+    :param save_db: Whether or not we want to save results to the database.
+        True by default.
+    :type save_db: bool
+    :param rewrite: Whether or not to rewrite existing files.  If true, it will
+        ignore files in Mongo and recreate them.  If false, it will download
+        files based on the latest available completed job.
+    :type rewrite: bool
     :param debug: If true, prints out params for each job & module.
     :type debug: bool
     """
 
-    def __init__(self, dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs, base_path, debug=False):
-        super(AlignGenerator, self).__init__(dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs, base_path, debug)
+    def __init__(self, dax, master_jobs, master_files, mdb, workflow_module,
+                workflow_jobs, base_path, save_db=True, rewrite=False, debug=False):
+        super(AlignGenerator, self).__init__(
+            dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs,
+            base_path, save_db=save_db, rewrite=rewrite, debug=debug
+        )
         self.generate_calls = {
             "bwa": {
                 1: self._bwa_single,
@@ -51,11 +62,10 @@ class AlignGenerator(ModuleGenerator):
         """
         base_file = run.genome.get_base_file()
         inputs = {"ref_genome": base_file["name"]}
-        additional_inputs = {}
         for f in run.genome.get_additional_files():
             ext = f["name"].split(".")[-1]
-            additional_inputs["ref_genome.%s" % (ext,)] = f["name"]
-        return (inputs, additional_inputs)
+            inputs["ref_genome.%s" % (ext,)] = f["name"]
+        return inputs
 
     def _bowtie2_init(self, run):
         """
@@ -72,12 +82,14 @@ class AlignGenerator(ModuleGenerator):
         """
         base_file = run.genome.get_base_file()
         additional_files = run.genome.get_additional_files()
-        inputs = {"ref_genome_prefix": run.genome.file_prefix}
-        additional_inputs = {"ref_genome": base_file["name"]}
+        inputs = {
+            "ref_genome_prefix": run.genome.file_prefix,
+            "ref_genome": base_file["name"]
+        }
         for f in run.genome.get_additional_files():
             ext = ".".join(f["name"].split(".")[-3 if "rev" in f["name"] else -2:])
-            additional_inputs["ref_genome.%s" % (ext,)] = f["name"]
-        return (inputs, additional_inputs)
+            inputs["ref_genome.%s" % (ext,)] = f["name"]
+        return inputs
 
     def _bwa_single(self, run, result):
         """
@@ -91,10 +103,10 @@ class AlignGenerator(ModuleGenerator):
         be processed
         """
         markers = {"tool": "bwa", "read_end": "single"}
-        inputs, additional_inputs = self._bwa_init(run)
+        inputs = self._bwa_init(run)
         prev_results = self.get_prev_results(run, result)
         inputs["download_1.fastq"] = prev_results[0].full_name
-        return (markers, inputs, additional_inputs)
+        return (markers, inputs)
 
     def _bwa_paired(self, run, result):
         """
@@ -107,11 +119,11 @@ class AlignGenerator(ModuleGenerator):
         jobs with paired end reads.
         """
         markers = {"tool": "bwa", "read_end": "paired"}
-        inputs, additional_inputs = self._bwa_init(run)
+        inputs = self._bwa_init(run)
         prev_results = self.get_prev_results(run, result)
         inputs["download_1.fastq"] = prev_results[0].full_name
         inputs["download_2.fastq"] = prev_results[1].full_name
-        return (markers, inputs, additional_inputs)
+        return (markers, inputs)
 
     def _bowtie2_single(self, run, result):
         """
@@ -125,10 +137,10 @@ class AlignGenerator(ModuleGenerator):
         be processed.
         """
         markers = {"tool": "bowtie2", "read_end": "single"}
-        inputs, additional_inputs = self._bowtie2_init(run)
+        inputs = self._bowtie2_init(run)
         prev_results = self.get_prev_results(run, result)
         inputs["download_1.fastq"] = prev_results[0].full_name
-        return (markers, inputs, additional_inputs)
+        return (markers, inputs)
 
 
     def _bowtie2_paired(self, run, result):
@@ -142,11 +154,11 @@ class AlignGenerator(ModuleGenerator):
         jobs with paired end reads.
         """
         markers = {"tool": "bowtie2", "read_end": "paired"}
-        inputs, additional_inputs = self._bowtie2_init(run)
+        inputs = self._bowtie2_init(run)
         prev_results = self.get_prev_results(run, result)
         inputs["download_1.fastq"] = prev_results[0].full_name
         inputs["download_2.fastq"] = prev_results[1].full_name
-        return (markers, inputs, additional_inputs)
+        return (markers, inputs)
 
     def _find_pairs(self, sample_list):
         """
@@ -236,6 +248,6 @@ class AlignGenerator(ModuleGenerator):
         :type result: :py:class:`~chipathlon.result.Result`
         """
         prev_results = self.get_prev_results(run, result)
-        markers, inputs, additional_inputs = self.generate_calls[run.genome.tool][len(prev_results)](run, result)
+        markers, inputs = self.generate_calls[run.genome.tool][len(prev_results)](run, result)
         results = self.create_results(run, result)
-        return markers, inputs, additional_inputs, self.get_outputs(results)
+        return markers, inputs, self.get_outputs(results)

chipathlon/generators/idr_generator.py

@@ -12,17 +12,28 @@ class IdrGenerator(ModuleGenerator):
     :param mdb: A MongoDB database class for fetching sample meta data.
     :type mdb: :py:class:`~chipathlon.db.MongoDB`
     :param workflow_module: The actual module being used.
-    :type workflow_module: :py:class:`~chipathlon.workflow_module.WorkflowModule`
+    :type workflow_module: `~chipathlon.workflow_module.WorkflowModule`
     :param workflow_jobs: Dictionary mapping workflow_job name -> workflow_job instance
     :type workflow_jobs: dict
     :param base_path: Base location of the workflow, used to save metadata files.
     :type base_path: str
+    :param save_db: Whether or not we want to save results to the database.
+        True by default.
+    :type save_db: bool
+    :param rewrite: Whether or not to rewrite existing files.  If true, it will
+        ignore files in Mongo and recreate them.  If false, it will download
+        files based on the latest available completed job.
+    :type rewrite: bool
     :param debug: If true, prints out params for each job & module.
     :type debug: bool
     """
 
-    def __init__(self, dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs, base_path, debug=False):
-        super(IdrGenerator, self).__init__(dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs, base_path, debug)
+    def __init__(self, dax, master_jobs, master_files, mdb, workflow_module,
+                workflow_jobs, base_path, save_db=True, rewrite=False, debug=False):
+        super(IdrGenerator, self).__init__(
+            dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs,
+            base_path, save_db=save_db, rewrite=rewrite, debug=debug
+        )
         self.module_name = "idr"
         self.result_dict = {}
         self.output_files = {
@@ -114,6 +125,5 @@ class IdrGenerator(ModuleGenerator):
             "sorted_sample_1.bed": result_pair[0].full_name,
             "sorted_sample_2.bed": result_pair[1].full_name
         }
-        additional_inputs = {}
         results = self.create_results(run, result)
-        return markers, inputs, additional_inputs, self.get_outputs(results)
+        return markers, inputs, self.get_outputs(results)

chipathlon/generators/module_generator.py

@@ -18,11 +18,19 @@ class ModuleGenerator(object):
     :type workflow_jobs: dict
     :param base_path: Base location of the workflow, used to save metadata files.
     :type base_path: str
+    :param save_db: Whether or not we want to save results to the database.
+        True by default.
+    :type save_db: bool
+    :param rewrite: Whether or not to rewrite existing files.  If true, it will
+        ignore files in Mongo and recreate them.  If false, it will download
+        files based on the latest available completed job.
+    :type rewrite: bool
     :param debug: If true, prints out params for each job & module.
     :type debug: bool
     """
 
-    def __init__(self, dax, master_jobs, master_files, mdb, workflow_module, workflow_jobs, base_path, debug=False):
+    def __init__(self, dax, master_jobs, master_files, mdb, workflow_module,
+                workflow_jobs, base_path, save_db=True, rewrite=False, debug=False):
         self.dax = dax
         self.master_jobs = master_jobs
         self.master_files = master_files
@@ -30,20 +38,20 @@ class ModuleGenerator(object):
         self.module = workflow_module
         self.workflow_jobs = workflow_jobs
         self.base_path = base_path
+        self.save_db = save_db
+        self.rewrite = rewrite
         self.debug = debug
         self.prev_results = {}
         self.save_results = []
         pprint.PrettyPrinter(indent=4)
         return
 
-    def _print_parse(self, run, markers, inputs, additional_inputs, outputs):
+    def _print_parse(self, run, markers, inputs, outputs):
         """
         :param markers: The splits to take in a given module.
         :type markers: dict
         :param inputs: The inputs required for a given module.
         :type inputs: dict
-        :param additional_inputs: The additional_inputs required for a given module.
-        :type additional_inputs: dict
         :param outputs: The outputs required for a given module.
         :type outputs: dict
         :param genome: The genome for the run
@@ -51,12 +59,11 @@ class ModuleGenerator(object):
 
         Prints the generated data from a single parse call in a pretty format.
         """
-        print "\n[%s]\n\tParsing Run: %s\n\tMarkers: %s\n\tInputs: %s\n\tAdditional: %s\n\tOutputs: %s" % (
+        print "\n[%s]\n\tParsing Run: %s\n\tMarkers: %s\n\tInputs: %s\n\tOutputs: %s" % (
             self.__class__.__name__,
             run,
             markers,
             inputs,
-            additional_inputs,
             outputs
         )
         return
@@ -75,16 +82,17 @@ class ModuleGenerator(object):
         all_jobs = prev_results[0].all_jobs
         all_markers = final_result.all_markers
         results = []
+        # Iterate through the jobs from the workflow module
         for job_dict in self.module.get_job_list(all_markers[self.module.name]):
             job_name = job_dict.keys()[0]
             job_info = job_dict[job_name]
             all_jobs.append(self.workflow_jobs[job_name])
             if job_info.get("outputs"):
-                for output_dict in job_info["outputs"]:
-                    logical_name = output_dict.keys()[0]
-                    output_info = output_dict[logical_name]
-                    # We have to explicitly compare since None evalutes to False.
+                for logical_name, output_info in job_info["outputs"].iteritems():
+                    # We have to explicitly compare since None evaluates to False.
                     if not output_info.get("final_result") == True:
+                        # Create the intermediate result with all jobs &
+                        # markers that have been run up to this point
                         result = Result(
                             logical_name,
                             final_result.control_samples,
@@ -95,6 +103,9 @@ class ModuleGenerator(object):
                             prefix_join=output_info.get("prefix_join"),
                             name_template=output_info.get("name_template")
                         )
+                        # Some results require additional prefixes i.e.
+                        # peakranger produces two separate output files
+                        # that we need to keep track of.
                         if final_result.get_meta("add_prefix") is not None:
                             result.add_to_prefix(final_result.get_meta("add_prefix"))
                         results.append(result)
@@ -102,7 +113,8 @@ class ModuleGenerator(object):
                     else:
                         result = run.find_result(self.module.name, logical_name, final_result)
                         results.append(result)
-                    if result.should_save:
+                    # Should we save the results to the db?
+                    if result.should_save and self.save_db:
                         self.save_results.append(result)
 
         return results
@@ -128,16 +140,32 @@ class ModuleGenerator(object):
         Download the target result from gridfs.
         """
         db_result = result.get_db_result(self.mdb, run.genome)
-        inputs = [
-            {"name": "host", "type": "string", "value": self.mdb.host},
-            {"name": "username", "type": "string", "value": self.mdb.username},
-            {"name": "password", "type": "string", "value": self.mdb.password},
-            {"name": "id", "type": "string", "value": db_result["gridfs_id"]}
-        ]
-        outputs = [
-            {"name": result.full_name, "type": "file", "file": result.pegasus_file, "transfer": False}
-        ]
-        download_job = self.workflow_jobs["download_from_gridfs"].create_job(inputs, [], outputs)
+        inputs = {
+            "host": {
+                "name": "host",
+                "value": self.mdb.host
+            },
+            "username": {
+                "name": "username",
+                "value": self.mdb.username
+            },
+            "password": {
+                "name": "password",
+                "value": self.mdb.password
+            },
+            "gfs_id": {
+                "name": "gfs_id",
+                "value": db_result["gridfs_id"]
+            }
+        }
+        outputs = {
+            "downloaded_result": {
+                "name": result.full_name,
+                "file": result.pegasus_file,
+                "transfer": False
+            }
+        }
+        download_job = self.workflow_jobs["download_from_gridfs"].create_job(inputs, outputs)
         if download_job is not None:
             self.master_files[result.full_name] = result.pegasus_file
             self.dax.addJob(download_job)
@@ -153,14 +181,24 @@ class ModuleGenerator(object):
         """
         # Assuming we only call on a result file with 1 accession
         sample = result.control_samples[0] if len(result.control_samples) > 0 else result.signal_samples[0]
-        inputs = [
-            {"name": "url", "type": "string", "value": sample["hcc_url"] if "hcc_url" in sample else sample["url"]},
-            {"name": "md5", "type": "string", "value": sample["md5sum"]}
-        ]
-        outputs = [
-            {"name": result.full_name, "type": "file", "file": result.pegasus_file, "transfer": False}
-        ]
-        download_job = self.workflow_jobs["download_from_encode"].create_job(inputs, [], outputs)
+        inputs = {
+            "url": {
+                "name": "url",
+                "value": sample["hcc_url"] if "hcc_url" in sample else sample["url"]
+            },
+            "md5": {
+                "name": "md5",
+                "value": sample["md5sum"]
+            }
+        }
+        outputs = {
+            "downloaded_file": {
+                "name": result.full_name,
+                "file": result.pegasus_file,
+                "transfer": False
+            }
+        }
+        download_job = self.workflow_jobs["download_from_encode"].create_job(inputs, outputs)
         if download_job is not None:
             self.master_files[result.full_name] = result.pegasus_file
             self.dax.addJob(download_job)
@@ -184,12 +222,10 @@ class ModuleGenerator(object):
                 self._download_from_gridfs(run, result)
         return
 
-    def generate(self, run, rewrite=False):
+    def generate(self, run):
         """
         :param run: The target run to generate jobs for.
         :type run: :py:class:`~chipathlon.run.Run`
-        :param rewrite: Whether or not to rewrite data even if it exists.
-        :type rewrite: bool
 
         Generates actual workflow jobs for a particular run.  The logic for
         job generation is based on the fact that all output files are unique,
@@ -210,14 +246,16 @@ class ModuleGenerator(object):
         """
         final_results = self.create_final_results(run)
         for result in final_results:
-            if rewrite or not result.exists_in_db(self.mdb, run.genome):
+            if self.rewrite or not result.exists_in_db(self.mdb, run.genome):
                 for prev_result in self.get_prev_results(run, result):
-                    if prev_result.exists_in_encode() or (not rewrite and prev_result.exists_in_db(self.mdb, run.genome)):
+                    if prev_result.exists_in_encode() or (not self.rewrite and prev_result.exists_in_db(self.mdb, run.genome)):
                         self.add_download_job(run, prev_result)
                 # We only want to add jobs for the very last result in a module
                 # Otherwise we will get duplicate jobs.
                 if result.last_result:
                     self.add_jobs(run, result)
+            else:
+                print "Skipping jobs for run: %s, result: %s" % (run, result)
         return
 
     def add_jobs(self, run, result):
@@ -230,10 +268,10 @@ class ModuleGenerator(object):
         This function is used to generate all jobs
         necessary to create a particular result file.
         """
-        markers, inputs, additional_inputs, outputs = self.parse_result(run, result)
+        markers, inputs, outputs = self.parse_result(run, result)
         if self.debug:
-            self._print_parse(run, markers, inputs, additional_inputs, outputs)
-        self.module.add_jobs(self.dax, self.master_jobs, self.master_files, markers, inputs, additional_inputs, outputs)
+            self._print_parse(run, markers, inputs, outputs)
+        self.module.add_jobs(self.dax, self.master_jobs, self.master_files, markers, inputs, outputs)
         while len(self.save_results) > 0:
             res = self.save_results.pop()
             res.save_to_db(self.mdb, self.dax, self.master_files, self.master_jobs, run.genome, self.workflow_jobs["db_save_result"], self.base_path)
@@ -248,7 +286,7 @@ class ModuleGenerator(object):
 
         This function should return everything necessary for
         actually creating jobs from the module i.e.
-        markers, inputs, additional_inputs, and outputs
+        markers, inputs, and outputs
 
         markers -- Should be a dictionary that has key value pairs for the
         current module splits that should be taken.  For example, the align
@@ -260,8 +298,6 @@ class ModuleGenerator(object):
         'align.bam' as input, so the inputs dictionary should be
         {"align.bam": "full_file_name.bam"}
 
-        additional_inputs -- Same as inputs, not directly used by the jobs.
-
         outputs -- Same as inputs except defines the outputs created by the
         jobs within the module.
         """

chipathlon/generators/peak_call_generator.py

@@ -13,17 +13,28 @@ class PeakCallGenerator(ModuleGenerator):
     :param mdb: A MongoDB database class for fetching sample meta data.
     :type mdb: :py:class:`~chipathlon.db.MongoDB`
     :param workflow_module: The actual module being used.
-    :type workflow_module: :py:class:`~chipathlon.workflow_module.WorkflowModule`
+    :type workflow_module: `~chipathlon.workflow_module.WorkflowModule`
     :param workflow_jobs: Dictionary mapping workflow_job name -> workflow_job instance
     :type workflow_jobs: dict
     :param base_path: Base location of the workflow, used to save metadata files.
     :type base_path: str
+    :param save_db: Whether or not we want to save results to the database.
+        True by default.
+    :type save_db: bool
+    :param rewrite: Whether or not to rewrite existing files.  If true, it will