def runcoasim(outputdir,
populationSpec,
NeRef,
L,
r,
outgroup,
migrationSpec=False):
out = open(outputdir + "/trees.dnd", "w")
if migrationSpec:
arg = CoaSim.simulate([],
populationSpec,
migrationSpec=migrationSpec,
rho=4 * NeRef * L * r,
keepEmptyIntervals=True)
else:
arg = CoaSim.simulate([],
populationSpec,
rho=4 * NeRef * L * r,
keepEmptyIntervals=True)
for i in arg.intervals:
s = str(i.tree)
if outgroup > 0:
s.replace(')', ' ')
s = s.replace(')', ' ')
s = s.replace('(', ' ')
s = s.replace("'0'", ' ')
s = s.replace("'1'", ' ')
s = s.replace("'2'", ' ')
s = s.replace(':', ' ')
s = s.replace(';', ' ')
s = s.replace(',', ' ')
s = s.replace(' ', ' ')
s = s.replace(' ', ' ')
s = s.replace(' ', ' ')
s = s.strip()
ss = s.split()
max = 0
for e in ss:
if float(e) > max:
max = float(e)
s = str(i.tree)
s = s.replace(';', '')
val = outgroup - max
s = '(' + s + ' : ' + str(val) + ", 'outgroup' : " + str(
outgroup) + ');'
out.writelines(str(i.start) + " " + str(i.end) + " " + s + "\n")
out.close()
#!/bin/env python
import CoaSim
from CoaSim.popStructure import Population as P, Merge as M, Sample as S, \
Migration as Mi
popSpec = P(1,M(1.5,[P(1,S(4),name='1'),P(1,S(4),name='2')]))
migSpec = [Mi('1','2',1),Mi('2','1',1)]
tree = CoaSim.simulate([], popSpec, migSpec,
keepEmptyIntervals=True,
keepMigrationEvents=True,
seed=10).intervals[0].tree
assert str(tree) == "(((('5' : 0.259312,('0' : 0.171585,'3' : 0.171585) : [&migration={1,0.218047,0}] 0.0877268) : 0.13414,('2' : 0.218047,'1' : 0.218047) : [&migration={1,0.218047,0}] 0.175404) : 0.491754,('7' : 0.0821288,'4' : 0.0821288) : 0.803077) : 0.117671,'6' : [&migration={1,0.205103,0,0.171585,1}] 1.00288);"
def simulateCoasim(spec, **args):
# test if all args are there
if "length" not in args:
raise Exception("Sequence 'length' parameter required")
if "NeRef" not in args:
raise Exception("'NeRef' parameter required")
if "r" not in args:
raise Exception("'r' (recombination rate?) parameter required")
if "g" not in args:
raise Exception("'g' (Generation time) parameter required")
if "u" not in args:
raise Exception("'u' (mutation rate) parameter required")
if "optionsFile" not in args:
raise Exception("'optionsFile' parameter required")
if "migration" not in args:
args["migration"] = []
# test T's and N's
count = 0
t = "T1"
while t in args:
count = count + 1
t = t + str(len(t))
# Test N
n = "N"
for i in range(1, count + 1):
if "N" + str(i) not in args:
raise Exception("'N" + str(i) + "' parameter required")
n = n + str(i)
if n not in args:
raise Exception("'" + n + "' parameter required")
# run the suff
st = spec(**args)
kei = True
if "keepEmptyIntervals" in args:
kei = args["keepEmptyIntervals"]
# print "simulating"
arg = CoaSim.simulate(
[],
st,
rho=4 * args["NeRef"] * args["length"] * args["r"],
keepEmptyIntervals=kei,
migrationSpec=args["migration"],
) # what about 4
if "recMap" in args:
# map intervals and parse trees
trees = mapIntervalsAndParseTrees(arg.intervals, args["recMap"])
else:
# parse trees
trees = []
for interval in arg.intervals:
trees.append((interval.start, interval.end, parse_tree(str(interval.tree))))
# reroot trees
if "addOutGroup" in args and args["addOutGroup"]:
trees = addOutGroup(trees, **args)
# scale trees
trees = scale_trees(trees, 2.0 * args["NeRef"] * args["g"] * args["u"])
if "tree_height_file" in args:
fout = open(args["tree_height_file"], "w")
bps = args["length"]
if "tree_visitor" not in args:
th = TreeHeight()
else:
th = args["tree_visitor"]
for (start, end, tree) in trees:
th.reset()
tree.dfs_traverse(th)
h = th.get_result()
to_write = "%s\n" % str(h)
for i in range(int(start * bps), int(end * bps)):
fout.write(to_write)
fout.close()
# for (s,e,t) in trees:
# print s,e,t
if "hook" in args:
args["hook"].run(arg, trees, args)
# args["hook"].parseTrees(arg, args)
return bppSeqGen(trees, **args)
def mergeEvent(self, pops, t, k):
self.counts['me'] += 1
from CoaSim.popStructure import Population as P, Merge as M, Sample as S, \
Bottleneck as B, Growth as G, \
Migration as Mi
cb = callbacks()
CoaSim.simulate(
[],
P(1,
M(1.5, [P(1, S(2), name='1'), P(1, S(2), name='2')]),
epochs=[B(.2, 1.5, 2), G(10, 2)]),
[Mi('1', '2', 0.001), Mi('2', '1', 0.002)],
rho=40,
Q=10,
gamma=2,
keepEmptyIntervals=True,
seed=10,
callbacks=cb)
# this is just regression testing, not proper testing :-(
expected = {
'me': 1,
'bn_leave': 1,
'mig': 4,
'g_enter': 1,
'coa': 173,
'gc': 10,
'bn_enter': 1,
#!/bin/env python
import CoaSim
markers = [CoaSim.SNPMarker(0.2, 0.1, 0.9),
CoaSim.SNPMarker(0.3, 0.1, 0.9)]
intervals = CoaSim.simulate(markers, 5, rho=4, seed=10).intervals
i1 = intervals[0]
t1 = i1.tree
r = t1.root
assert r.isAncestral(i1.start)
assert not r.isAncestral(i1.end)
assert r.isAncestral(i1.start + (i1.end-i1.start)/2)
assert not r.isAncestral(i1.start - 1e-10)
assert r.isAncestral(i1.end - 1e-10)
assert r.children == r.children # testing comparison...
# regression testing...
assert len(r.children) == 2
assert len(r.children[0].children) == 1
#!/bin/env python
import CoaSim
markers = [CoaSim.SNPMarker(0.2, 0.1, 0.9), CoaSim.SNPMarker(0.3, 0.1, 0.9)]
intervals = CoaSim.simulate(markers, 5, rho=4, seed=10).intervals
trees = [i.tree for i in intervals]
assert intervals == [t.interval for t in trees]
# regression testing... FIXME: dependent on local random number generator...
assert [t.branchLength
for t in trees] == [6.8694553479721083, 4.8932502412522147]
assert [t.height for t in trees] == [2.0935405459452552, 1.2403594539408294]
assert str(
t
) == "(('3' : 0.970517,'0' : 0.970517) : 0.269843,(('4' : 0.638217,'2' : 0.638217) : 0.165581,'1' : 0.803798) : 0.436561);"
#!/bin/env python
import CoaSim
markers = [CoaSim.SNPMarker(0.2, 0.1, 0.9),
CoaSim.SNPMarker(0.3, 0.1, 0.9)]
# first a very simple regression test ... not really thorough
seqs = CoaSim.simulate(markers, 5, rho=40, seed=10).sequences
assert seqs == [[0, 0], [0, 0], [1, 0], [0, 0], [0, 1]]
# test incorrect input
try:
CoaSim.simulate([markers[1],markers[0]], 5)
assert False
except ValueError, e:
assert str(e) == 'Marker positions out of sequence.'
try:
CoaSim.simulate(markers, -5)
assert False
except ValueError, e:
assert str(e) == 'Non-positive sample size.'
try:
CoaSim.simulate(markers, 5, rho=-2)
assert False
except ValueError, e:
assert str(e) == 'Negative rate or intensity: -2.'
#!/bin/env python
import CoaSim
arg = CoaSim.simulate([],
5,
rho=20,
gamma=10,
Q=2,
seed=10,
keepEmptyIntervals=True)
counter = dict()
for node in arg.nodes:
try:
counter[type(node)] += 1
except KeyError:
counter[type(node)] = 1
assert len(counter) == 4
assert counter[CoaSim.LeafNode] == 5
assert counter[CoaSim.CoalescentNode] == 72
assert counter[CoaSim.RecombinationNode] == 88
assert counter[CoaSim.GeneConversionNode] == 82
from CoaSim.popStructure import Population as P, Merge as M, Sample as S, \
Bottleneck as B, Growth as G, \
Migration as Mi
#!/bin/env python
import CoaSim
markers = [CoaSim.SNPMarker(0.2, 0.1, 0.9),
CoaSim.SNPMarker(0.3, 0.1, 0.9)]
assert len(CoaSim.simulate(markers, 5).intervals) == 1 # no recomb, 1 interval
arg = CoaSim.simulate(markers, 5, rho=4)
assert arg.intervals == arg.intervals # check that comparison works
# this wont work, however: assert arg.intervals[0] is arg.intervals[0]
# since we get new objects for each call to intervals.
# regression testing... FIXME: dependent on local random number generator...
arg = CoaSim.simulate(markers, 5, rho=4, seed=10)
assert [i.start for i in arg.intervals] == [0.092338229572557953, 0.2567803758460937]
assert [i.end for i in arg.intervals] == [0.2458365472247063, 0.42215607660922971]
assert [i.length for i in arg.intervals] == [0.15349831765214833, 0.16537570076313601]
#!/bin/env python
import CoaSim
markers = [CoaSim.SNPMarker(0.2, 0.1, 0.9), CoaSim.SNPMarker(0.3, 0.1, 0.9)]
assert len(CoaSim.simulate(markers, 5).intervals) == 1 # no recomb, 1 interval
arg = CoaSim.simulate(markers, 5, rho=4)
assert arg.intervals == arg.intervals # check that comparison works
# this wont work, however: assert arg.intervals[0] is arg.intervals[0]
# since we get new objects for each call to intervals.
# regression testing... FIXME: dependent on local random number generator...
arg = CoaSim.simulate(markers, 5, rho=4, seed=10)
assert [i.start
for i in arg.intervals] == [0.092338229572557953, 0.2567803758460937]
assert [i.end
for i in arg.intervals] == [0.2458365472247063, 0.42215607660922971]
assert [i.length
for i in arg.intervals] == [0.15349831765214833, 0.16537570076313601]
#!/bin/env python
import CoaSim
arg = CoaSim.simulate([], 5, rho=20, gamma=10, Q=2,
seed=10, keepEmptyIntervals=True)
counter = dict()
for node in arg.nodes:
try:
counter[type(node)] += 1
except KeyError:
counter[type(node)] = 1
assert len(counter) == 4
assert counter[CoaSim.LeafNode] == 5
assert counter[CoaSim.CoalescentNode] == 72
assert counter[CoaSim.RecombinationNode] == 88
assert counter[CoaSim.GeneConversionNode] == 82
from CoaSim.popStructure import Population as P, Merge as M, Sample as S, \
Bottleneck as B, Growth as G, \
Migration as Mi
arg = CoaSim.simulate([], P(1,M(1.5,[P(1,S(2),name='1'),P(1,S(2),name='2')])),
[Mi('1','2',0.001),Mi('2','1',0.002)],
keepEmptyIntervals=True, seed=10)
#!/bin/env python
import CoaSim
from CoaSim.popStructure import Population as P, Merge as M, Sample as S, \
Migration as Mi
popSpec = P(1, M(1.5, [P(1, S(4), name='1'), P(1, S(4), name='2')]))
migSpec = [Mi('1', '2', 1), Mi('2', '1', 1)]
tree = CoaSim.simulate([],
popSpec,
migSpec,
keepEmptyIntervals=True,
keepMigrationEvents=True,
seed=10).intervals[0].tree
assert str(
tree
) == "(((('5' : 0.259312,('0' : 0.171585,'3' : 0.171585) : [&migration={1,0.218047,0}] 0.0877268) : 0.13414,('2' : 0.218047,'1' : 0.218047) : [&migration={1,0.218047,0}] 0.175404) : 0.491754,('7' : 0.0821288,'4' : 0.0821288) : 0.803077) : 0.117671,'6' : [&migration={1,0.205103,0,0.171585,1}] 1.00288);"
def simulateCoasim(spec, **args):
# test if all args are there
if "length" not in args:
raise Exception("Sequence 'length' parameter required")
if "NeRef" not in args:
raise Exception("'NeRef' parameter required")
if "r" not in args:
raise Exception("'r' (recombination rate?) parameter required")
if "g" not in args:
raise Exception("'g' (Generation time) parameter required")
if "u" not in args:
raise Exception("'u' (mutation rate) parameter required")
if "optionsFile" not in args:
raise Exception("'optionsFile' parameter required")
if "migration" not in args:
args["migration"] = []
# test T's and N's
count = 0
t = "T1"
while t in args:
count = count + 1
t = t + str(len(t))
# Test N
n = "N"
for i in range(1, count + 1):
if "N" + str(i) not in args:
raise Exception("'N" + str(i) + "' parameter required")
n = n + str(i)
if n not in args:
raise Exception("'" + n + "' parameter required")
# run the stuff
st = spec(**args)
kei = True
if "keepEmptyIntervals" in args:
kei = args["keepEmptyIntervals"]
# print "simulating"
arg = CoaSim.simulate(
[],
st,
rho=4 * args["NeRef"] * args["length"] * args["r"],
keepEmptyIntervals=kei,
migrationSpec=args["migration"],
) # what about 4
## # FIXME: possibly do this instead: collapse all runs of identical trees:
## trees = list()
## prevTree = None
## for tree in [(x.start, x.end, str(x.tree)) for x in arg.intervals]:
## if tree == prevTree:
## trees[-1][1] = end
## else:
## trees.append([start, end, tree])
## prevTree = tree
## if "recMap" in args:
## # map intervals and parse trees
## trees = mapIntervalsAndParseTrees(trees, args["recMap"])
## else:
## # parse trees
## for i in range(len(trees)):
## trees[i][2] = parse_tree(trees[i][2])
if "recMap" in args:
# map intervals and parse trees
trees = mapIntervalsAndParseTrees([(x.start, x.end, x.tree) for x in arg.intervals], args["recMap"])
else:
# parse trees
trees = []
for interval in arg.intervals:
trees.append((interval.start, interval.end, parse_tree(str(interval.tree))))
# reroot trees
if "addOutGroup" in args and args["addOutGroup"]:
trees = addOutGroup(trees, **args)
# scale trees
trees = scale_trees(trees, 2.0 * args["NeRef"] * args["g"] * args["u"])
if "tree_height_file" in args:
fout = open(args["tree_height_file"], "w")
bps = args["length"]
if "tree_visitor" not in args:
th = TreeHeight()
else:
th = args["tree_visitor"]
for (start, end, tree) in trees:
th.reset()
tree.dfs_traverse(th)
h = th.get_result()
to_write = "%s\n" % str(h)
for i in range(int(start * bps), int(end * bps)):
fout.write(to_write)
fout.close()
# for (s,e,t) in trees:
# print s,e,t
if "hook" in args:
args["hook"].run(arg, trees, args)
# args["hook"].parseTrees(arg, args)
# for t in trees:
# print t
# ####
# def getTree(tree):
# left, right = tree.get_edges()
# if isinstance(left[0], Leaf) and left[0].identifier == '3': # mayby identifier is not an int but a str...
# # outgroup is included - call on ingroup
# return right[0]
# elif isinstance(right[0], Leaf) and right[0].identifier == '3': # mayby identifier is not an int but a str...
# # outgroup is included - call on ingroup
# return left[0]
# else:
# return tree
#
# th = TreeHeight()
#
# for (start, end, tree) in trees:
# th.reset()
# t = getTree(tree)
# t.dfs_traverse(th)
# h = th.get_result()
# print h
#
#
# with open("coasim.trees", 'w') as f:
# for s, e, t in trees:
# print >>f, s, e, t
# sys.exit()
####
return bppSeqGen(trees, **args)
def simulateCoasim(spec, **args):
#test if all args are there
if "length" not in args:
raise Exception("Sequence 'length' parameter required")
if "NeRef" not in args:
raise Exception("'NeRef' parameter required")
if "r" not in args:
raise Exception("'r' (recombination rate?) parameter required")
if "g" not in args:
raise Exception("'g' (Generation time) parameter required")
if "u" not in args:
raise Exception("'u' (mutation rate) parameter required")
if "optionsFile" not in args:
raise Exception("'optionsFile' parameter required")
if "migration" not in args:
args["migration"] = []
#test T's and N's
count = 0;
t = "T1"
while t in args:
count = count +1
t = t + str(len(t))
#Test N
n = "N"
for i in range(1, count+1):
if "N" + str(i) not in args:
raise Exception("'N" + str(i) + "' parameter required")
n = n + str(i)
if n not in args:
raise Exception("'" + n + "' parameter required")
#run the suff
st = spec(**args)
kei = True
if "keepEmptyIntervals" in args:
kei = args["keepEmptyIntervals"];
# print "simulating"
arg = CoaSim.simulate([], st, rho = 4 * args["NeRef"] * args["length"] * args["r"],
keepEmptyIntervals=kei, migrationSpec=args["migration"]) #what about 4
trees = []
if "hotspotWidth" in args and "hotspotRatio" in args:
# hotspotWidth
hw = args["hotspotWidth"]
# ratio of hotspot non-hotstop recombination rate:
F = args["hotspotRatio"]
p = hw / (1 - hw)
# width of hotspot region before contraction:
ohw = (F*p)/(F*p+1)
# expansion factor:
e = (1-hw) / (1-ohw)
# contraction factor:
c = hw / ohw
args["hotSpotBackgroundRate"] = args["r"] * (1-hw)/(1-ohw)
print >>sys.stderr, "Simulating hotspot", args["hotSpotBackgroundRate"]
# # nr hotspots:
# nrHotSpots = 4
# # o iohw into
# m = generateMap(nrHotSpots, ohw, c, e):
# map
m = [(0, e), ((1-ohw)/2.0, c), (((1-ohw)/2.0)+ohw, e), (1.0, None)]
j = 0
prevEnd = 0
recPoints = list()
for i, interval in enumerate(arg.intervals):
while j < len(m)-1 and interval.start >= m[j+1][0]:
j += 1
start = prevEnd
if j < len(m)-1 and interval.end > m[j+1][0]:
end = prevEnd + (m[j+1][0] - interval.start) * m[j][1] + (interval.end - m[j+1][0]) * m[j+1][1]
else:
end = prevEnd + (interval.end - interval.start) * m[j][1]
prevEnd = end
trees.append((start, end, parse_tree(str(interval.tree))))
else:
for interval in arg.intervals:
trees.append((interval.start, interval.end, parse_tree(str(interval.tree))))
# for s, e, t in trees:
# print s, e
# import sys
# sys.exit()
#reroot trees
if "addOutGroup" in args and args["addOutGroup"]:
trees = addOutGroup(trees, **args)
# print "scaling"
#scale trees
trees = scale_trees(trees, 2.0*args["NeRef"]*args["g"]*args["u"])
if "tree_height_file" in args:
fout = open(args["tree_height_file"], "w")
bps = args["length"]
if "tree_visitor" not in args:
th = TreeHeight()
else:
th = args["tree_visitor"]
for (start, end, tree) in trees:
th.reset();
tree.dfs_traverse(th)
h = th.get_result()
to_write = "%s\n" % str(h)
for i in range(int(start*bps), int(end*bps)):
fout.write(to_write)
fout.close()
#for (s,e,t) in trees:
#print s,e,t
if "hook" in args:
args["hook"].parseARG(arg, trees, args)
# args["hook"].parseTrees(arg, args)
return bppSeqGen(trees, **args)
def migrationEvent(self, pop1, pop2, t, k):
self.counts['mig'] += 1
def mergeEvent(self, pops, t, k):
self.counts['me'] += 1
from CoaSim.popStructure import Population as P, Merge as M, Sample as S, \
Bottleneck as B, Growth as G, \
Migration as Mi
cb = callbacks()
CoaSim.simulate([], P(1,M(1.5,[P(1,S(2),name='1'),P(1,S(2),name='2')]),
epochs=[B(.2,1.5,2),G(10,2)]),
[Mi('1','2',0.001),Mi('2','1',0.002)],
rho=40,Q=10,gamma=2,
keepEmptyIntervals=True, seed=10,
callbacks=cb)
# this is just regression testing, not proper testing :-(
expected = {'me': 1, 'bn_leave': 1, 'mig': 4, 'g_enter': 1, 'coa': 173,
'gc': 10, 'bn_enter': 1, 'g_leave': 0, 'rec': 168}
assert cb.counts == expected
class AllDone: pass
class ex_callbacks(callbacks):
def coalescentEvent(self, n, k):
callbacks.coalescentEvent(self, n, k)
if self.counts['coa'] == 120:
raise AllDone
#!/bin/env python
import CoaSim
markers = [CoaSim.SNPMarker(0.2, 0.1, 0.9), CoaSim.SNPMarker(0.3, 0.1, 0.9)]
# first a very simple regression test ... not really thorough
seqs = CoaSim.simulate(markers, 5, rho=40, seed=10).sequences
assert seqs == [[0, 0], [0, 0], [1, 0], [0, 0], [0, 1]]
# test incorrect input
try:
CoaSim.simulate([markers[1], markers[0]], 5)
assert False
except ValueError, e:
assert str(e) == 'Marker positions out of sequence.'
try:
CoaSim.simulate(markers, -5)
assert False
except ValueError, e:
assert str(e) == 'Non-positive sample size.'
try:
CoaSim.simulate(markers, 5, rho=-2)
assert False
except ValueError, e:
assert str(e) == 'Negative rate or intensity: -2.'
try:
CoaSim.simulate(markers, 5, gamma=-2)
