def main(ddir, rrange = None):
read_parameters(ddir)
tf = open("%s/tree" % (ddir))
t = newickToTree(tf.read())
root = "N1"
parent, children = findParentsAndChildren(t, root)
treestr = rootedTreeToStr(parent, children, root)
open("%s/tree.full" % (ddir), "w").write(treestr)
evidence_cpds = read_evidence_cpds("%s/%s" % (ddir, common.CPD_DIR))
internal_cpds = read_internal_cpds("%s/%s" % (ddir, common.TREE_CPD_FILE))
ef = open("%s/evidence" % (ddir))
evidence = read_evidence(ef)
model, observed_species, ancestors = build_model(t, evidence_cpds, internal_cpds)
naive_model, spec2, anc2 = build_naive_model(t, evidence_cpds)
blast_model, spec3, anc3 = build_blast_model(t, evidence_cpds, internal_cpds)
gtg_model, spec4, anc4 = build_gtg_model(t, evidence_cpds, internal_cpds)
complete_evidence(evidence, observed_species)
# compute posteriors
of = open("%s/reaction-scores" % (ddir), "w")
of.write("#Generated by \"%s\" on %s\n" % (" ".join(sys.argv), datetime.datetime.now()))
if rrange != None:
of.write("#Reaction range: %s\n" % (rrange))
compute_reaction_scores(model, naive_model,
blast_model, gtg_model, evidence,
observed_species, ancestors, of, rrange)
def main(ddir, rrange=None):
read_parameters(ddir)
tf = open("%s/tree" % (ddir))
t = newickToTree(tf.read())
root = "N1"
parent, children = findParentsAndChildren(t, root)
treestr = rootedTreeToStr(parent, children, root)
open("%s/tree.full" % (ddir), "w").write(treestr)
evidence_cpds = read_evidence_cpds("%s/%s" % (ddir, common.CPD_DIR))
internal_cpds = read_internal_cpds("%s/%s" % (ddir, common.TREE_CPD_FILE))
ef = open("%s/evidence" % (ddir))
evidence = read_evidence(ef)
model, observed_species, ancestors = build_model(t, evidence_cpds,
internal_cpds)
naive_model, spec2, anc2 = build_naive_model(t, evidence_cpds)
blast_model, spec3, anc3 = build_blast_model(t, evidence_cpds,
internal_cpds)
gtg_model, spec4, anc4 = build_gtg_model(t, evidence_cpds, internal_cpds)
complete_evidence(evidence, observed_species)
# compute posteriors
of = open("%s/reaction-scores" % (ddir), "w")
of.write("#Generated by \"%s\" on %s\n" %
(" ".join(sys.argv), datetime.datetime.now()))
if rrange != None:
of.write("#Reaction range: %s\n" % (rrange))
compute_reaction_scores(model, naive_model, blast_model, gtg_model,
evidence, observed_species, ancestors, of, rrange)
def build_dag(T, use_tree = True, blast = True, gtg = True):
parents, children = findParentsAndChildren(T, ROOT_NODE)
dag = Digraph()
observed_species = set()
ancestors = set()
if use_tree:
for u in children:
ancestors.add(u)
for v in children[u]:
dag.add_edge(u, v)
for u in T:
if u not in children:
observed_species.add(u)
if blast:
dag.add_edge(u, get_blast_node_name(u))
if gtg:
dag.add_edge(u, get_gtg_node_name(u))
return dag, observed_species, ancestors
def build_dag(T, use_tree=True, blast=True, gtg=True):
parents, children = findParentsAndChildren(T, ROOT_NODE)
dag = Digraph()
observed_species = set()
ancestors = set()
if use_tree:
for u in children:
ancestors.add(u)
for v in children[u]:
dag.add_edge(u, v)
for u in T:
if u not in children:
observed_species.add(u)
if blast:
dag.add_edge(u, get_blast_node_name(u))
if gtg:
dag.add_edge(u, get_gtg_node_name(u))
return dag, observed_species, ancestors
def estimate(mdir, species_tree_fn, odir):
of = open("%s/%s" % (odir, common.TREE_CPD_FILE), "w")
models, all_ecs = common.read_models(mdir)
t = common.read_tree(species_tree_fn)
parents, children = tree.findParentsAndChildren(t, "N1")
snadd, sndel, snnoadd, snnodel = 0, 0, 0, 0
all_mutations = {}
for i in range(500):
nadd, ndel, nnoadd, nnodel, mutations = fitch.parsimony(
parents, children, models, all_ecs)
snadd += nadd
sndel += ndel
snnoadd += nnoadd
snnodel += nnodel
for key in mutations:
if key not in all_mutations:
all_mutations[key] = [0, 0, 0, 0]
all_mutations[key][0] += mutations[key][0] + 1
all_mutations[key][1] += mutations[key][1] + 1
all_mutations[key][2] += mutations[key][2] + 1
all_mutations[key][3] += mutations[key][3] + 1
#print all_mutations
keys = all_mutations.keys()
keys.sort()
for key in keys:
nadd, ndel, nnadd, nndel = all_mutations[key]
padd = 1.0 * nadd / (nadd + nnadd)
pdel = 1.0 * ndel / (ndel + nndel)
u, v = key.split("-")
of.write("%s\t%s\t%s\t%s\n" % (u, v, padd, pdel))
#print key, padd, pdel, nadd, ndel, nnadd, nndel
#print "p-add\t%s\t%s" % (1.0 * snadd / (snadd + snnoadd), snadd)
#print "p-del\t%s\t%s" % (1.0 * sndel / (sndel + snnodel), sndel)
#print "p-no-add\t%s\t%s" % (1.0 * snnoadd / (snadd + snnoadd), snnoadd)
#print "p-no-del\t%s\t%s" % (1.0 * snnodel / (sndel + snnodel), snnodel)
of.close()
def estimate(mdir, species_tree_fn, odir):
of = open("%s/%s" % (odir, common.TREE_CPD_FILE), "w")
models, all_ecs = common.read_models(mdir)
t = common.read_tree(species_tree_fn)
parents, children = tree.findParentsAndChildren(t, "N1")
snadd, sndel, snnoadd, snnodel = 0, 0, 0, 0
all_mutations = {}
for i in range(500):
nadd, ndel, nnoadd, nnodel, mutations = fitch.parsimony(parents, children, models, all_ecs)
snadd += nadd
sndel += ndel
snnoadd += nnoadd
snnodel += nnodel
for key in mutations:
if key not in all_mutations:
all_mutations[key] = [0, 0, 0, 0]
all_mutations[key][0] += mutations[key][0] + 1
all_mutations[key][1] += mutations[key][1] + 1
all_mutations[key][2] += mutations[key][2] + 1
all_mutations[key][3] += mutations[key][3] + 1
#print all_mutations
keys = all_mutations.keys()
keys.sort()
for key in keys:
nadd, ndel, nnadd, nndel = all_mutations[key]
padd = 1.0 * nadd / (nadd + nnadd)
pdel = 1.0 * ndel / (ndel + nndel)
u, v = key.split("-")
of.write("%s\t%s\t%s\t%s\n" % (u, v, padd, pdel))
#print key, padd, pdel, nadd, ndel, nnadd, nndel
#print "p-add\t%s\t%s" % (1.0 * snadd / (snadd + snnoadd), snadd)
#print "p-del\t%s\t%s" % (1.0 * sndel / (sndel + snnodel), sndel)
#print "p-no-add\t%s\t%s" % (1.0 * snnoadd / (snadd + snnoadd), snnoadd)
#print "p-no-del\t%s\t%s" % (1.0 * snnodel / (sndel + snnodel), snnodel)
of.close()
