def from_graph_peaks_in_fasta(cls, graph, vg_graph_json_file_name,
chromosome, fasta_file_name,
regions_bed_file, true_peaks_file):
reads = PeakCollection.from_fasta_file(fasta_file_name, graph=graph)
vg_graph = pyvg.vg.Graph.create_from_file(
vg_graph_json_file_name, limit_to_chromosomes=chromosome)
logging.info("Finding linear path")
linear_path_file = "linear_path_%s.intervalcollection" % chromosome
try:
linear_path = obg.IntervalCollection.from_file(
linear_path_file, text_file=True).intervals[0]
linear_path = linear_path.to_indexed_interval()
except FileNotFoundError:
linear_path = create_linear_path(graph,
vg_graph,
path_name=chromosome,
write_to_file=linear_path_file)
linear_path.graph = graph
filtered_reads = []
# Convert regions to intervals in graph
logging.info("Converting regions to regions in graph")
graph_regions = []
bed_file = open(regions_bed_file)
for line in bed_file:
print(line)
line = line.split()
chr = line[0]
start = int(line[1])
end = int(line[2])
if chr != "chr%s" % chromosome:
logging.info("Skipping %s, %d, %d" % (chr, start, end))
continue
graph_interval = linear_path.get_subinterval(start, end)
graph_regions.append(graph_interval)
assert len(graph_regions
) > 0, " Found not graph regions for chr %d" % chromosome
graph_regions = PeakCollection(graph_regions)
# Filter out reads not overlapping with linear regions
for read in reads:
n_overlapping = graph_regions.get_overlapping_intervals(
read, minimum_overlap=1)
if n_overlapping:
filtered_reads.append(read)
logging.info("Found %d reads in graph regions" % len(filtered_reads))
return cls(chromosome, reads, true_peaks_file)
def macs_to_graph_peaks(folder):
for chrom in ["1", "2", "3", "4", "5"]:
path = NumpyIndexedInterval.from_file("/data/bioinf/tair2/" + chrom +
"_linear_pathv2.interval")
graph = Graph.from_file("/data/bioinf/tair2/" + chrom + ".nobg")
macs_peaks = PeakCollection.from_fasta_file(
folder + "/macs_sequences_chr%s_summits_unique.fasta" % chrom,
graph)
macs_peaks.to_file(
folder +
"/%s_macs_unique_graph_summits.intervalcollection" % chrom, True)
def test_intervals_to_fasta_from_fasta(self):
run_argument_parser([
"create_ob_graph", "-o", "tests/testgraph.obg",
"tests/vg_test_graph.json"
])
PeakCollection([Peak(0, 2, [1, 2], score=3)
]).to_file("tests/testintervals.intervalcollection",
text_file=True)
run_argument_parser([
"peaks_to_fasta", "tests/testgraph.obg.sequences",
"tests/testintervals.intervalcollection",
"tests/testsequences.fasta"
])
collection = PeakCollection.from_fasta_file(
"tests/testsequences.fasta")
self.assertEqual(len(collection.intervals), 1)
self.assertEqual(collection.intervals[0].sequence.lower(), "tttcccctt")
def test_get_summits(self):
qvalues = SparseValues(np.array([0]), np.array([3]))
qvalues.track_size = 22
qvalues.to_sparse_files("tests/test_qvalues")
run_argument_parser([
"create_ob_graph", "-o", "tests/testgraph.obg",
"tests/vg_test_graph.json"
])
max_paths = PeakCollection([Peak(0, 2, [1, 2], score=3)])
PeakFasta(self.correct_sequence_graph).write_max_path_sequences(
"tests/test_max_paths.fasta", max_paths)
run_argument_parser([
"get_summits", "-g", "tests/testgraph.obg",
"tests/test_max_paths.fasta", "tests/test_qvalues", "2"
])
result = PeakCollection.from_fasta_file(
"tests/test_max_paths_summits.fasta")
self.assertEqual(result.intervals[0], Peak(2, 6, [1]))
self.assertEqual(result.intervals[0].sequence.lower(), "tccc")
return mapping
out_file = open("motif_summary_graph_matching_macs.tsv", "w")
for chrom in ["1", "2", "3", "4", "5"]:
logging.info("Chromosome %s" % chrom)
path = NumpyIndexedInterval.from_file("/data/bioinf/tair2/" + chrom +
"_linear_pathv2.interval")
graph = Graph.from_file("/data/bioinf/tair2/" + chrom + ".nobg")
macs_peaks = NonGraphPeakCollection.from_fasta("macs_sequences_chr" +
chrom + "_summits.fasta")
macs_peaks = PeakCollection.create_from_nongraph_peak_collection(
graph, macs_peaks, path)
macs_peaks.create_node_index()
graph_peaks = PeakCollection.from_fasta_file(chrom +
"_sequences_summits.fasta")
graph_peaks.create_node_index()
macs_motif_matches = set([
line.split("\t")[2]
for line in open("fimo_macs_chr" + chrom + "/fimo.txt")
if not line.startswith("#")
])
graph_motif_matches = set([
line.split("\t")[2]
for line in open("fimo_graph_chr" + chrom + "/fimo.txt")
if not line.startswith("#")
])
mapping = get_id_mapping(graph_peaks, macs_peaks)
motif_ids = [