],
defaults={'annotations_dir': ""})
PARAMS = P.PARAMS
PARAMS_ANNOTATIONS = P.peek_parameters(PARAMS["annotations_dir"], "genesets")
###################################################################
###################################################################
###################################################################
# Helper functions mapping tracks to conditions, etc
###################################################################
# load all tracks - exclude input/control tracks
Sample = tracks.Sample
TRACKS = tracks.Tracks(Sample).loadFromDirectory(glob.glob("*.bed.gz"),
"(\S+).bed.gz")
TRACKS_BEDFILES = ["%s.bed.gz" % x for x in TRACKS]
def getAssociatedBAMFiles(track):
'''return a list of BAM files associated with a track.
By default, this method searches for ``track.bam`` file in the
current directory and returns an offset of 0.
Associations can be defined in the .yml file in the section
[bams]. For example, the following snippet associates track
track1 with the bamfiles :file:`track1.bam` and :file:`track2.bam`::
[bams]
###################################################################
###################################################################
###################################################################
# Helper functions mapping tracks to conditions, etc
###################################################################
# load all tracks - exclude input/control tracks
# determine the location of the input files (reads).
DATADIR = PARAMS.get('input', '.')
if not os.path.exists(DATADIR):
raise OSError('data directory %s does not exists')
Sample = PipelineTracks.Sample
TRACKS = PipelineTracks.Tracks(Sample).loadFromDirectory(
glob.glob(os.path.join(DATADIR, "*.bed.gz")),
"(\S+).bed.gz")
BEDFILES = [os.path.join(
DATADIR, "%s.bed.gz") % x for x in TRACKS]
# create an indicator target
@transform(BEDFILES, suffix(".gz"), ".gz")
def BedFiles(infile, outfile):
pass
BAMFILES = glob.glob(os.path.join(DATADIR, "*.bam"))
dbh = sqlite3.connect(PARAMS["database_name"])
statement = '''ATTACH DATABASE '%s' as annotations''' % (
PARAMS["annotations_database"])
cc = dbh.cursor()
cc.execute(statement)
cc.close()
return dbh
class MySample(tracks.Sample):
attributes = tuple(PARAMS["attributes"].split(","))
TRACKS = tracks.Tracks(MySample).loadFromDirectory(glob.glob("*.bam"),
"(\S+).bam")
Sample = tracks.AutoSample
DESIGNS = tracks.Tracks(Sample).loadFromDirectory(glob.glob("*.design.tsv"),
"(\S+).design.tsv")
###################################################################
###################################################################
###################################################################
# DEXSeq workflow
###################################################################
@mkdir("results.dir")
@files(PARAMS["annotations_interface_geneset_all_gtf"], "geneset_flat.gff")
def buildGff(infile, outfile):
try:
PARAMS["input"]
except KeyError:
DATADIR = "."
else:
if PARAMS["input"] == 0:
DATADIR = "."
elif PARAMS["input"] == 1:
DATADIR = "data.dir"
else:
DATADIR = PARAMS["input"] # not recommended practise.
Sample = tracks.AutoSample
# collect sra and fastq.gz tracks
BAM_TRACKS = tracks.Tracks(Sample).loadFromDirectory(glob.glob("*.bam"),
"(\S+).bam")
DESIGNS = tracks.Tracks(Sample).loadFromDirectory(glob.glob("design*.tsv"),
"design(\S+).tsv")
# do not use - legacy methods
# here only to stop ruffus erroring. Remove once pipleine scrum is
# complete and old code has been removed
GENESETS = tracks.Tracks(Sample).loadFromDirectory(glob.glob("*.gtf.gz"),
"(\S+).gtf.gz")
TRACKS = tracks.Tracks(Sample).loadFromDirectory(glob.glob("*.bam"),
"(\S+).bam")
# group by experiment (assume that last field is a replicate identifier)
EXPERIMENTS = tracks.Aggregate(BAM_TRACKS, labels=("condition", "tissue"))
###############################################################################