def test_get_bulge_dimensions(self):
bg = fgb.BulgeGraph(dotbracket_str='(.(.))')
bd = bg.get_bulge_dimensions('i0')
self.assertEquals(bd, (1,0))
bg = fgb.BulgeGraph(dotbracket_str='((.).)')
bd = bg.get_bulge_dimensions('i0')
self.assertEquals(bd, (0,1))
bg = fgb.BulgeGraph(dotbracket_str='().()')
bd = bg.get_bulge_dimensions('m0')
dotbracket = '(.(.).(.).(.))'
bg = fgb.BulgeGraph(dotbracket_str=dotbracket)
bd = bg.get_bulge_dimensions('m0')
self.assertEquals(bd, (1,1000))
bd = bg.get_bulge_dimensions('m1')
self.assertEquals(bd, (0,1000))
bd = bg.get_bulge_dimensions('m2')
self.assertEquals(bd, (1,1000))
bd = bg.get_bulge_dimensions('m3')
self.assertEquals(bd, (1,1000))
bg = fgb.BulgeGraph(dotbracket_str='((..((..))....))..((..((..))...))')
bd = bg.get_bulge_dimensions('i0')
self.assertEquals(bd, (2, 4))
bd = bg.get_bulge_dimensions('i1')
self.assertEquals(bd, (2, 3))
def test_get_flanking_handles(self):
bg = fgb.BulgeGraph(dotbracket_str='((..))')
h = bg.get_flanking_handles('h0')
self.assertEqual(h, (2, 5, 1, 4))
bg = fgb.BulgeGraph(dotbracket_str='((.((.)).(.).))')
self.assertEqual(bg.get_flanking_handles('m0'),
(2,4,1,3))
self.assertEqual(bg.get_flanking_handles('m2'),
(8,10,1,3))
self.assertEqual(bg.get_flanking_handles('m1'),
(12,14,0,2))
bg = fgb.BulgeGraph(dotbracket_str='(.(.).).(.(.))')
self.assertEqual(bg.get_flanking_handles('i0', side=0),
(1,3,0,2))
self.assertEqual(bg.get_flanking_handles('i0', side=1),
(5,7,0,2))
self.assertEqual(bg.get_flanking_handles('i1', side=0),
(9,11,0,2))
self.assertEqual(bg.get_flanking_handles('i1', side=1),
(13,14,0,1))
bg = fgb.BulgeGraph(dotbracket_str='((.((.)).)).((.((.))))')
# 1234567890123456789012
self.assertEqual(bg.get_flanking_handles('i0', side=0),
(2,4,1,3))
self.assertEqual(bg.get_flanking_handles('i0', side=1),
(8,10,1,3))
self.assertEqual(bg.get_flanking_handles('i1', side=0),
(14,16,1,3))
self.assertEqual(bg.get_flanking_handles('i1', side=1),
(20,21,1,2))
def test_define_residue_num_iterator(self):
bg = fgb.BulgeGraph(dotbracket_str='((..((..))((..))))')
drni = bg.define_residue_num_iterator('m2', adjacent=True)
# the second multiloop should have at least two adjacent nucleotides
self.assertEqual(len(list(drni)), 2)
drni = bg.define_residue_num_iterator('m1', adjacent=True)
# the second multiloop should have at least two adjacent nucleotides
self.assertEqual(len(list(drni)), 2)
drni = bg.define_residue_num_iterator('m1', adjacent=True)
bg = fgb.BulgeGraph()
bg.from_dotbracket('..((..((...))..))..((..))..')
self.assertEqual(list(bg.define_residue_num_iterator('f1')),
[1,2])
self.assertEqual(list(bg.define_residue_num_iterator('t1')),
[26, 27])
self.assertEqual(list(bg.define_residue_num_iterator('s1')),
[7, 8, 12, 13])
self.assertEqual(list(bg.define_residue_num_iterator('i0')),
[5,6,14,15])
fa=""">blah
AAAAAAAAAA
((((.)).))
"""
bg.from_fasta(fa, dissolve_length_one_stems=True)
self.assertEqual(list(bg.define_residue_num_iterator('i0', adjacent=True)),
[2,3,7,8,9])
self.assertEqual(list(bg.define_residue_num_iterator('i0', adjacent=True, seq_ids=True)),
[(' ', 2, ' '), (' ', 3, ' '), (' ', 7, ' '), (' ', 8, ' '), (' ', 9, ' ')])
def test_from_dotplot(self):
bg = fgb.BulgeGraph()
bg.from_dotbracket(self.dotbracket)
self.assertEquals(bg.seq_length, len(self.dotbracket))
bg = fgb.BulgeGraph()
bg.from_dotbracket('....')
def test_get_any_sides(self):
bg = fgb.BulgeGraph(dotbracket_str='((..((..))..)).((..))')
self.assertEqual(bg.get_any_sides('s0', 'i0'), (1,0))
self.assertEqual(bg.get_any_sides('i0', 's0'), (0,1))
bg = fgb.BulgeGraph(dotbracket_str='((..((..))((..))))')
self.assertEqual(bg.get_any_sides('s1', 'm1'), (0, 1))
self.assertEqual(bg.get_any_sides('m1', 's1'), (1, 0))
def annotate_structures(input_file, output_file):
""" Annotate secondary structure predictions with structural contexts.
Given dot-bracket strings this function will annote every character
as either 'H' (hairpin), 'S' (stem), 'I' (internal loop/bulge) or 'M' (multi loop). The input file
must be a fasta formatted file and each sequence and structure must span a single line:
'>header
'CCCCAUAGGGG
'((((...)))) (-3.3)
This is the default format of e.g. RNAfold. The output file will contain the annotated string:
'>header
'CCCCAUAGGGG
'SSSSHHHSSSS
Parameters
----------
input_file : str
A fasta file containing secondary structure predictions.
output_file : str
A fasta file with secondary structure annotations.
"""
handle_in = get_handle(input_file, "rt")
handle_out = get_handle(output_file, "wt")
for header, entry in parse_fasta(handle_in, "_"):
entry = entry.split("_")
bg = cgb.BulgeGraph()
bg.from_dotbracket(entry[1].split()[0])
handle_out.write(">{}\n".format(header))
handle_out.write("{}\n{}\n".format(entry[0], bg.to_element_string().upper()))
handle_in.close()
handle_out.close()
def removeGUWobble(seq1, seq2):
with open('dummyAl_structureCalc2way4wayStoreData_v2_20190416.txt',
'w') as f:
processCall = subprocess.Popen(
'RNAalifold --noPS --noGU --temp=20 dummyClustal_structureCalc2way4wayStoreData_v2_20190416.aln',
shell=True,
stdout=f,
stdin=subprocess.PIPE
) #temperature had to be lowered from default of 37C as otherwise some AT rich structures were not folding--note that this does change the folding for some RNAs as well
time.sleep(
1
) #important otherwise dummyAl_structureCalc2way4way_v1_20190222.txt not created before being accessed
dotBracket = returnDotBracket(
'dummyAl_structureCalc2way4wayStoreData_v2_20190416.txt')
bg = fgb.BulgeGraph(dotbracket_str=dotBracket)
listDotBracket = list(dotBracket)
for eachStem in bg.stem_iterator():
for eachBp in bg.stem_bp_iterator(
eachStem): #returns 1-index based numbering
i1 = eachBp[0] - 1
i2 = eachBp[1] - 1
if isGU(i1, i2, seq1) and isGU(i1, i2, seq2):
listDotBracket[i1] = '.'
listDotBracket[i2] = '.'
return ''.join(listDotBracket)
def translateIntoContexts(dotBracketString):
bg = fgb.BulgeGraph()
bg.from_dotbracket(dotBracketString)
rawContextString = upper(bg.to_element_string())
contextString1 = rawContextString.replace('F', 'E')
contextString = contextString1.replace('T', 'E')
return contextString
def main():
usage = """
python bpseq_to_bulge_graph.py secondary_structure.bpseq
"""
num_args= 1
parser = OptionParser(usage=usage)
#parser.add_option('-o', '--options', dest='some_option', default='yo', help="Place holder for a real option", type='str')
#parser.add_option('-u', '--useless', dest='uselesss', default=False, action='store_true', help='Another useless option')
(options, args) = parser.parse_args()
if len(args) < num_args:
parser.print_help()
sys.exit(1)
with open(args[0], 'r') as f:
text = f.read()
try:
int(text[0])
except ValueError:
i=text.find("\n1 ")
text=text[i+1:]
bg = fgb.BulgeGraph()
bg.from_bpseq_str(text)
print bg.to_bg_string()
def main(seq):
start = time.clock()
s1 = readInSeq(seq)
length = len(s1)
global structure
structure = [None for i in range(0, length)]
W = [[None for c in range(1,length+3)] for d in range(1,length+3)]
V = [[None for c in range(1,length+3)] for d in range(1,length+3)]
initW, initV = initialize(W, V, length)
calcW, calcV, maxScore = fillMatrix(initW, initV, s1, length)
traceWpath(0, length-1, calcW, calcV)
# note: the value of 'structure' has been changed to what you want
annot = ''.join(['.' if j is None else ')' if j < i else '(' for i, j in enumerate(structure)])
# getScoreRNA(tracedStruct, length)
# Annotate dot-bracket notation
bg = fgb.BulgeGraph()
bg.from_dotbracket(annot)
# print structure
print s1
print "The structure in string form: " + annot
print "The corresponding annotated notation: " + bg.to_element_string()
print "\nFinished in ",time.clock() - start,"seconds"
return (calcW, calcV)
def test_get_sides(self):
with open('test/forgi/data/1ymo.bpseq', 'r') as f:
lines = f.readlines()
bpseq_str = "".join(lines)
bg = fgb.BulgeGraph()
bg.from_bpseq_str(bpseq_str, dissolve_length_one_stems=True)
def main():
usage = """
./longest_stem.py dotbracket_file
"""
num_args = 1
parser = OptionParser()
#parser.add_option('-o', '--options', dest='some_option', default='yo', help="Place holder for a real option", type='str')
#parser.add_option('-u', '--useless', dest='uselesss', default=False, action='store_true', help='Another useless option')
(options, args) = parser.parse_args()
if len(args) < num_args:
parser.print_help()
sys.exit(1)
if args[0] == '-':
f = sys.stdin
else:
f = open(args[0])
brackets = "".join(f.readlines()).replace('\n', '')
bg = cgb.BulgeGraph()
bg.from_dotbracket(brackets)
biggest_stem = (-1, 'x')
for s in bg.stem_iterator():
if bg.stem_length(s) > biggest_stem[0]:
biggest_stem = (bg.stem_length(s), s)
print(biggest_stem[0])
def test_from_dotplot3(self):
dotbracket = '(.(.((((((...((((((....((((.((((.(((..(((((((((....)))))))))..((.......))....)))......))))))))...))))))..)).))))).)..((((..((((((((((...))))))))).))))).......'
bg = fgb.BulgeGraph()
self.check_graph_integrity(bg)
bg.from_dotbracket(dotbracket)
self.check_graph_integrity(bg)
def test_get_length(self):
bg = fgb.BulgeGraph(dotbracket_str='(())')
bg = fgb.BulgeGraph(dotbracket_str='((..))..(((.)))')
self.assertEquals(bg.get_length('s0'), 2)
self.assertEquals(bg.get_length('h0'), 2)
self.assertEquals(bg.get_length('m0'), 2)
self.assertEquals(bg.get_length('s1'), 3)
bg = fgb.BulgeGraph(dotbracket_str='(())(())')
self.assertEquals(bg.get_length('m0'), 0)
bg = fgb.BulgeGraph(dotbracket_str='(((((((((..(((..((((.(((((((((.....(((((.(((((....((((....))))....))))).....(((((((((.......)))))))))....))))).((........))...)))))))))))))...)))..))....))))))).')
self.assertEqual(bg.get_length('i4'), 2)
def test_find_multiloop_loops(self):
bg = fgb.BulgeGraph()
bg.from_dotbracket('((..((..))..((..))..))')
bg.find_multiloop_loops()
bg.from_dotbracket('((..((..((..))..((..))..))..((..))..))')
bg.from_dotbracket('(.(.(.(.).(.).).(.).))')
def test_to_bg_string(self):
self.fasta = """>1y26
CGCUUCAUAUAAUCCUAAUGAUAUGGUUUGGGAGUUUCUACCAAGAGCCUUAAACUCUUGAUUAUGAAGUG
(((((((((...((((((.........))))))........((((((.......))))))..)))))))))
"""
bg = fgb.BulgeGraph()
bg.from_fasta(self.fasta, dissolve_length_one_stems=True)
def test_stem_length(self):
bg = fgb.BulgeGraph(dotbracket_str='.((..(((..))).))((..))')
self.assertEqual(bg.stem_length('s0'), 2)
self.assertEqual(bg.stem_length('s1'), 3)
self.assertEqual(bg.stem_length('m0'), 0)
self.assertEqual(bg.stem_length('i0'), 1)
self.assertEqual(bg.stem_length('f1'), 1)
def test_connection_type(self):
bg = fgb.BulgeGraph(dotbracket_str='(.(.).).(.(.))')
self.assertEqual(bg.connection_type('m0', ['s0', 's2']), 3)
self.assertEqual(bg.connection_type('m0', ['s2', 's0']), -3)
self.assertEqual(bg.connection_type('i0', ['s0', 's1']), 1)
self.assertEqual(bg.connection_type('i0', ['s1', 's0']), -1)
def test_pairing_partner(self):
# documented
bg = fgb.BulgeGraph()
bg.from_dotbracket('((..))')
self.assertEquals(bg.pairing_partner(1), 6)
self.assertEquals(bg.pairing_partner(2), 5)
self.assertEquals(bg.pairing_partner(5), 2)
def test_remove_pseudoknots(self):
pk_fasta = '>hi\nAAAAAAAAAAAAAAAA\n((..[[[..))..]]]'
bg = fgb.BulgeGraph()
bg.from_fasta(pk_fasta)
dissolved_bp = forna.remove_pseudoknots(bg)
self.assertTrue(dissolved_bp is not None)
def get_element_with_id_list_from_dotbracket(dot_bracket_str):
graph = cgb.BulgeGraph()
graph.from_dotbracket(dot_bracket_str)
element_str = graph.to_element_string(True)
list_result = element_str.split('\n')
elements = list_result[0]
ids = list_result[1]
return zip([str(elem) for elem in elements], [str(elem_id) for elem_id in ids])
def test_are_adjacent_stems(self):
bg = fgb.BulgeGraph(dotbracket_str='((..((..))..))..((..))')
self.assertTrue(bg.are_adjacent_stems('s0', 's1'))
self.assertTrue(bg.are_adjacent_stems('s0', 's2'))
self.assertFalse(bg.are_adjacent_stems('s1', 's2'))
self.assertFalse(bg.are_adjacent_stems('s0', 's2',
multiloops_count=False))
def test_dissolve_stem(self):
'''
Test to make sure length one stems can be dissolved.
'''
bg = fgb.BulgeGraph()
bg.from_dotbracket('((.(..((..))..).))', dissolve_length_one_stems = True)
self.assertEquals(bg.to_dotbracket_string(), '((....((..))....))')
self.check_graph_integrity(bg)
bg = fgb.BulgeGraph(dotbracket_str='((..))..((..))')
self.assertEquals(bg.to_dotbracket_string(), '((..))..((..))')
bg.dissolve_stem('s0')
self.check_graph_integrity(bg)
self.assertEquals(bg.to_dotbracket_string(), '........((..))')
bg.dissolve_stem('s0')
self.check_graph_integrity(bg)
def test_from_fasta(self):
bg = fgb.BulgeGraph()
with open('test/forgi/threedee/data/3V2F.fa', 'r') as f:
text = f.read()
bg.from_fasta(text, dissolve_length_one_stems=False)
for s in bg.stem_iterator():
bg.stem_length(s)
def test_create_mst(self):
'''
Test the creation of a minimum spanning tree from the graph.
'''
db = '....((((((...((((((.....(((.((((.(((..(((((((((....)))))))))..((.......))....)))......)))))))....))))))..)).)))).....((((...(((((((((...)))))))))..)))).......'
bg = fgb.BulgeGraph(dotbracket_str=db)
mst = bg.get_mst()
self.assertTrue("m0" in mst)
build_order = bg.traverse_graph()
db = '..((.(())..(())...)).'
bg = fgb.BulgeGraph(dotbracket_str=db)
mst = bg.get_mst()
self.assertTrue('m0' in mst)
self.assertTrue('m2' in mst)
build_order = bg.traverse_graph()
def returnElementNotation(sequence):
with open('dummy_structureCalc2way4way_v1_20190404.txt', 'w') as f:
processCall = subprocess.Popen(['RNAfold', '--noPS', '--noGU'],
stdout=f,
stdin=subprocess.PIPE)
processCall.communicate(input=sequence)
dotBracket = returnDotBracket(
'dummy_structureCalc2way4way_v1_20190404.txt')
bg = fgb.BulgeGraph(dotbracket_str=dotBracket)
return bg.to_element_string()
def stemThere(dotBracket):
bg = fgb.BulgeGraph(dotbracket_str=dotBracket)
elementRep = bg.to_element_string()
numberStems = 0
for eachEl in bg.stem_iterator():
numberStems += 1
if numberStems == 0:
return False
else:
return True
def test_random_subgraph(self):
bg = fgb.BulgeGraph(dotbracket_str='(.(.).).(.(.))..((..((..((..))..))..))')
sg = bg.random_subgraph()
# check to make sure there are no duplicate elements
self.assertEquals(len(sg), len(set(sg)))
nbg = fgb.bg_from_subgraph(bg, sg)
self.assertTrue(set(nbg.defines.keys()) == set(sg))
def test_get_define_seq_str(self):
bg = fgb.BulgeGraph(dotbracket_str="(.(.))")
bg.seq = 'acgauu'
self.assertEquals(bg.get_define_seq_str("i0"), ['c', ''])
bg = fgb.BulgeGraph(dotbracket_str="(.(.))")
bg.seq = 'acgauu'
self.assertEquals(bg.get_define_seq_str("i0", True), ['acg','uu'])
bg = fgb.BulgeGraph(dotbracket_str='(.(.).(.).)')
bg.seq = 'acguaaccggu'
self.assertEquals(bg.get_define_seq_str('m0'), ['c'])
self.assertEquals(bg.get_define_seq_str('m0', True), ['acg'])
self.assertEquals(bg.get_define_seq_str('m1'), ['g'])
self.assertEquals(bg.get_define_seq_str('m1', True), ['ggu'])
self.assertEquals(bg.get_define_seq_str('m2'), ['a'])
self.assertEquals(bg.get_define_seq_str('m2', True), ['aac'])
def json_to_json(rna_json_str):
'''
Convert an RNA json string to fasta file, then to a bulge_graph
and then back to a json.
The purpose is to maintain the integrity of the molecule and to
maintain the positions of all the hidden nodes after modification.
'''
with open('test.out', 'w') as f:
f.write(rna_json_str)
(all_fastas, all_xs, all_ys, all_uids,
different_tree_links) = json_to_fasta(rna_json_str)
big_json = {'nodes': [], 'links': []}
coords_to_index = dict()
for fasta_text, xs, ys, uids in zip(all_fastas, all_xs, all_ys, all_uids):
bg = fgb.BulgeGraph()
bg.from_fasta(fasta_text)
new_json = bg_to_json(bg, xs=xs, ys=ys, uids=uids)
for l in new_json['links']:
# the indices of the new nodes will be offset, so the links
# have to have their node pointers adjusted as well
l['source'] += len(big_json['nodes'])
l['target'] += len(big_json['nodes'])
big_json['links'] += [l]
# Create a mapping between the coordinates of a node and its index
# in the node list. To be used when creating links between different
# molecules, which are stored according to the coordinates of the nodes
# being linked
for i, n in enumerate(new_json['nodes']):
if n['node_type'] == 'nucleotide':
coords_to_index[(n['x'], n['y'])] = i + len(big_json['nodes'])
big_json['nodes'] += new_json['nodes']
# add the links that are between different molecules
for dtl in different_tree_links:
fud.pv('dtl')
n1 = coords_to_index[(dtl[0])]
n2 = coords_to_index[(dtl[1])]
fud.pv('n1,n2')
big_json['links'] += [{
'source': n1,
'target': n2,
'link_type': 'basepair',
'value': 1
}]
#fud.pv('big_json["nodes"]')
return big_json