# description of this example.
#
import simuPOP as sim
from simuPOP.utils import saveCSV
pop = sim.Population(size=[10],
loci=[2, 3],
lociNames=['r11', 'r12', 'r21', 'r22', 'r23'],
alleleNames=['A', 'B'],
infoFields='age')
sim.initSex(pop)
sim.initInfo(pop, [2, 3, 4], infoFields='age')
sim.initGenotype(pop, freq=[0.4, 0.6])
sim.maPenetrance(pop, loci=0, penetrance=(0.2, 0.2, 0.4))
# no filename so output to standard output
saveCSV(pop, infoFields='age')
# change affection code and how to output genotype
saveCSV(pop,
infoFields='age',
affectionFormatter={
True: 1,
False: 2
},
genoFormatter={
(0, 0): 'AA',
(0, 1): 'AB',
(1, 0): 'AB',
(1, 1): 'BB'
})
# save to a file
saveCSV(pop,
#Recombination
ops=[sim.Recombinator(rates=0.002)]),
postOps=[
#Mutation rate 10e-6
sim.SNPMutator(u=0.000001, v=0.000001)
],
#Evolve for a number 'numgen' of generations
gen=numgen)
#Getting population informations (number of subpopulations, population size)
sim.stat(pop, popSize=True)
subsize = pop.dvars().subPopSize
numpop = len(subsize)
#Setting environmental value for all individuals in each subpopulation
for i in range(numpop):
pop.setIndInfo(vec_env[i], 'env', subPop=i)
#Sampling 20 individuals at random in each population
sample = drawRandomSample(pop, sizes=[20] * numpop)
#Adding population name to the field of individuals
sample.addInfoFields('pop_name')
vecname = []
for i in range(1, numpop + 1):
vecname = vecname + [i] * 20
sample.setIndInfo(vecname, 'pop_name')
#Saving the data into a .csv format
saveCSV(sample,
filename="sim" + str(k) + ".csv",
infoFields=['pop_name', 'env'],
sexFormatter=None,
affectionFormatter=None,
header=False)
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:], shortopts="vhs:n:l:e:f:", longopts=["verbose", "help", "size=",
"number=", "loci=", "effect=",
"filename="])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
has_filename = False
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print ("Verbose mode")
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-s", "--size"):
individuals = o[1]
if verbose:
print "Population size is set at", individuals
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
has_filename = True
if verbose:
print "File will be saved as:", filename
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=int(individuals), loci=int(number), infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(), sim.InitGenotype(prop=[0.7, 0.3])], matingScheme=sim.RandomMating(),
postOps=[sim.PyQuanTrait(loci=loci, func=trait, infoFields=["qtrait"])],
gen=5)
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
if has_filename is False:
filename = "my"
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), phenotypes)))
f.close()
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def main():
# Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:], shortopts="vhp1:p2:s:n:l:e:f:i:m:g:r:", longopts=["verbose", "help", "parameter1=", "parameter2=", "size=",
"number=", "loci=", "effect=", "mean=", "filename=", "heritability=", "gen=", "rrate="])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
filename = "my"
size = 1000
p1 = aalpha = 2
p2 = bbeta = 3
number = 100
heritability = 0.2
mean = 2.0
gen = 5
rrate = 0.0
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print ("Verbose mode")
for o in opts:
if o[0] in ("-p1", "--parameter1"):
aalpha = float(o[1])
if verbose:
print "Gamma distribution will occur with alpha:", aalpha
for o in opts:
if o[0] in ("-p2", "--parameter2"):
bbeta = float(o[1])
if verbose:
print "Gamma distribution will occur with beta:", bbeta
for o in opts:
if o[0] in ("-s", "--size"):
individuals = o[1].split(",")
individuals = map(int, individuals)
if verbose:
print "Population size/s is set at", individuals
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
global loci
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
if verbose:
print "File will be saved as:", filename
elif o[0] in ("-i", "--heritability"):
heritability = float(o[1])
if verbose:
print "Heritability for simulation specified as:", heritability
elif o[0] in ("-m", "--mean"):
mean = o[1].split(",")
mean = map(float, mean)
if len(mean) == 1 and len(individuals) > 1:
mean = numpy.array(mean)
mean = numpy.repeat(mean, len(individuals), axis=0)
mean = list(mean)
if verbose:
print "Population mean/s specified as:", mean
elif o[0] in ("-g", "--gen"):
gen = int(o[1])
if verbose:
print "Generations to evolve specified as:", gen
elif o[0] in ("-r", "--rrate"):
rrate = float(o[1])
if verbose:
print "Recombination will occur with rate:", rrate
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=individuals, loci=int(number), infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(), sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(ops=sim.Recombinator(rates=rrate)),
postOps=[sim.PyQuanTrait(loci=loci, func=additive_model, infoFields=["qtrait"])],
gen=gen)
if verbose:
print "Coalescent process complete. Population evolved with", pop.numSubPop(),"sub-populations."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
# fun() obtains the heritability equation set to zero for various settings of sigma (standard deviation)
#NOTE: May need to tweak gamma distribution parameters to be appropriate for data!
def fun(sigma, h):
x_exact = list()
count = 0
for i in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_exact.append(current_mean + i)
count += 1
x_random = list()
#bbeta=((sigma**2)/current_mean) #Set up approximate beta variable for gamma distribution
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_random.append(random.normalvariate(current_mean + each, sigma))
count += 1
r = pearsonr(x_exact, x_random)[0]
return r - math.sqrt(h)
if verbose:
print "Building polynomial model for variance tuning..."
# Fits a polynomial model in numpy to the values obtained from the fun() function
points = list()
for i in drange(0, max(effects)*10, 0.001):
points.append(i)
y_points = list()
for i in points:
y_points.append(fun(i, heritability))
z = numpy.polyfit(x=points, y=y_points, deg=3)
p = numpy.poly1d(z)
# Netwon's method finds the polynomial model's roots
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations...\nPlease choose different settings."
usage()
sys.exit()
count += 1
xn = xn - p(xn)/p_d(xn)
if xn < 0.0:
xn = 0.0
if verbose:
print "Estimated variance of phenotypes for specified heriability: ", xn
return xn
if verbose:
print "Using Newton's method to find polynomial roots..."
# Files are saved to the specified location
estimated_variance = newton(p)
new_phenotypes = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
new_phenotypes.append(random.gammavariate((current_mean + each)/bbeta, ((estimated_variance/aalpha)**0.5)))
count += 1
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), new_phenotypes)))
f.close()
numpy.savetxt(filename + "_kt_ote2.txt", numpy.column_stack((loci, numpy.array(effects))))
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def qtrait(geno):
trait = random.normalvariate(sum(geno) * 5, random.uniform(0.0001, 3))
if trait <= 0:
trait = random.uniform(0.0001, 1)
return trait
pop.evolve(initOps=[sim.InitSex(),
sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(),
postOps=[
sim.PyQuanTrait(loci=(0, 1, 2, 3, 4, 5, 6, 10, 100),
func=qtrait,
infoFields=["qtrait"])
],
gen=10)
geno = list()
for i in pop.individuals():
geno.append(i.genotype())
pheno = list()
for i in pop.individuals():
pheno.append(i.qtrait)
f = open("qtrait1.txt", "w")
f.write("\n".join(map(lambda x: str(x), pheno)))
f.close()
saveCSV(pop, "sample.csv")
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:],
shortopts="vhs:n:l:e:f:i:",
longopts=[
"verbose", "help", "size=", "number=",
"loci=", "effect=", "filename=",
"herit="
])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
has_filename = False
has_heritability = False
filename = "my"
heritability = 0.2
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print("Verbose mode")
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-s", "--size"):
individuals = o[1]
if verbose:
print "Population size is set at", individuals
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
global loci
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
has_filename = True
if verbose:
print "File will be saved as:", filename
elif o[0] in ("-i", "--herit"):
heritability = float(o[1])
has_heritability = True
if verbose:
print "Heritability for simulation specified as:", heritability
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=int(individuals),
loci=int(number),
infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(),
sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(),
postOps=[
sim.PyQuanTrait(loci=loci,
func=additive_model,
infoFields=["qtrait"])
],
gen=5)
if verbose:
print "Coalescent process complete. Population evolved."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
#print i.genotype()
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
#print i.qtrait
def fun(sigma, h):
x_exact = phenotypes
x_random = list()
for each in phenotypes:
x_random.append(random.normalvariate(each, sigma))
r = pearsonr(x_exact, x_random)[0]
return r - math.sqrt(h)
#print fun(2.25, 0.25)
if verbose:
print "Building polynomial model for variance tuning..."
points = list()
for i in drange(0, max(effects) * 10, 0.001):
points.append(i)
y_points = list()
for i in points:
y_points.append(fun(i, heritability))
z = numpy.polyfit(x=points, y=y_points, deg=3)
p = numpy.poly1d(z)
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations..."
sys.exit()
count += 1
xn = xn - p(xn) / p_d(xn)
if verbose:
print "Estimated variance of phenotypes for specified heriability using Newton's method: ", xn
return xn
if verbose:
print "Using Newton's method to find polynomial roots..."
estimated_variance = newton(p)
new_phenotypes = list()
for each in phenotypes:
new_phenotypes.append(random.normalvariate(each, estimated_variance))
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), new_phenotypes)))
f.close()
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:],
shortopts="vhs:n:l:e:f:i:",
longopts=[
"verbose", "help", "size=", "number=",
"loci=", "effect=", "filename=",
"herit="
])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
has_filename = False
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print("Verbose mode")
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-s", "--size"):
individuals = o[1]
if verbose:
print "Population size is set at", individuals
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
global loci
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
has_filename = True
if verbose:
print "File will be saved as:", filename
elif o[0] in ("-i", "--herit"):
heritability = float(o[1])
has_heritability = True
if verbose:
print "Heritability for simulation specified as:", heritability
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=int(individuals),
loci=int(number),
infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(
initOps=[sim.InitSex(),
sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(),
#postOps=[sim.PyQuanTrait(loci=loci, func=trait, infoFields=["qtrait"])],
gen=5)
if verbose:
print "Population evolved."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
#print i.genotype()
def fun(sigma2, h):
exact_traits = list()
for i in genotypes:
exact_traits.append(exact_trait(i))
prob_traits = list()
for i in genotypes:
prob_traits.append(prob_trait(i, sigma2))
r = pearsonr(exact_traits, prob_traits)[0]
return r - math.sqrt(h)
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations..."
sys.exit()
count += 1
xn = xn - p(xn) / p_d(xn)
if verbose:
print "Estimated variance using Newton's method for solving roots is: ", xn
return xn
## Make sure to change to 100 points around the average "effects"
my_points = list()
for i in range(100):
my_points.append(fun(i, heritability))
z = numpy.polyfit(x=my_points, y=range(100), deg=2)
#print z
p = numpy.poly1d(z)
if verbose:
print "Polynomial fit for finding tuning parameter to match heritability: \n", p
estimated_variance = newton(p)
#print estimated_variance
phenotypes = list()
for i in pop.individuals():
phenotypes.append(prob_trait(i.genotype(), estimated_variance))
#print phenotypes
if has_filename is False:
filename = "my"
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), phenotypes)))
f.close()
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
output='>>pedigree4.ped'), #outputs pedigree file for checking
sim.Recombinator(rates=recom_map)
],
subPopSize=15000),
gen=1)
#save
#saveCSV(example_pop, filename='test_simuPOP_all.csv') #output all individuals (15,000 in this case)
#To subsample (this samples 1000 random individuals from last generation)
sub_sample = sim.sampling.drawRandomSample(example_pop, 1000)
#create array from lineage tracker with index labels for sampleid, locus, and allele for exporting
lin = np.array(sub_sample.lineage())
sample = sub_sample.indInfo('ind_id')
locus = founder.snpID.astype(str)
allele = ['a1', 'a2']
index = pd.MultiIndex.from_product([sample, allele, locus],
names=['sample', 'allele', 'snpID'])
lin2 = pd.DataFrame(data=lin, index=index, columns=['parent_of_origin'])
#Save lineage tracking information
lin2.to_csv('known_parent_of_origin.csv')
### Save 1000 sampled individuals, contains ind_id and 2*numLoci columns (1 for each allele)
saveCSV(sub_sample,
filename='known_GT.csv',
infoFields=['ind_id'],
sexFormatter=None,
affectionFormatter=None)
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:],
shortopts="vhs:n:l:e:f:",
longopts=[
"verbose", "help", "size=", "number=",
"loci=", "effect=", "filename="
])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
has_filename = False
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print("Verbose mode")
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-s", "--size"):
individuals = o[1]
if verbose:
print "Population size is set at", individuals
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
has_filename = True
if verbose:
print "File will be saved as:", filename
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=int(individuals),
loci=int(number),
infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(),
sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(),
postOps=[
sim.PyQuanTrait(loci=loci,
func=trait,
infoFields=["qtrait"])
],
gen=5)
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
if has_filename is False:
filename = "my"
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), phenotypes)))
f.close()
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:], shortopts="vhs:n:l:e:f:i:m:g:r:", longopts=["verbose", "help", "size=",
"number=", "loci=", "effect=", "mean=", "filename=", "heritability=", "gen=", "rrate="])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
filename = "my"
size = 1000
number = 100
heritability = 0.2
mean = 2.0
gen = 5
rrate = 0.0
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print ("Verbose mode")
for o in opts:
if o[0] in ("-s", "--size"):
individuals = o[1].split(",")
individuals = map(int, individuals)
if verbose:
print "Population size/s is set at", individuals
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
global loci
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
if verbose:
print "File will be saved as:", filename
elif o[0] in ("-i", "--heritability"):
heritability = float(o[1])
if verbose:
print "Heritability for simulation specified as:", heritability
elif o[0] in ("-m", "--mean"):
mean = o[1].split(",")
mean = map(float, mean)
if len(mean) == 1 and len(individuals) > 1:
mean = numpy.array(mean)
mean = numpy.repeat(mean, len(individuals), axis=0)
mean = list(mean)
if verbose:
print "Population mean/s specified as:", mean
elif o[0] in ("-g", "--gen"):
gen = int(o[1])
if verbose:
print "Generations to evolve specified as:", gen
elif o[0] in ("-r", "--rrate"):
rrate = float(o[1])
if verbose:
print "Recombination will occur with rate:", rrate
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=individuals, loci=int(number), infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(), sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(ops=sim.Recombinator(rates=rrate)),
postOps=[sim.PyQuanTrait(loci=loci, func=additive_model, infoFields=["qtrait"])],
gen=gen)
if verbose:
print "Coalescent process complete. Population evolved with", pop.numSubPop(),"sub-populations."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
def fun(sigma, h):
x_exact = list()
count = 0
for i in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_exact.append(current_mean + i)
count += 1
x_random = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_random.append(random.normalvariate(current_mean + each, sigma))
count += 1
r = pearsonr(x_exact, x_random)[0]
return r - math.sqrt(h)
if verbose:
print "Building polynomial model for variance tuning..."
points = list()
for i in drange(0, max(effects)*10, 0.001):
points.append(i)
y_points = list()
for i in points:
y_points.append(fun(i, heritability))
z = numpy.polyfit(x=points, y=y_points, deg=3)
p = numpy.poly1d(z)
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations...\nPlease choose different settings."
usage()
sys.exit()
count += 1
xn = xn - p(xn)/p_d(xn)
if xn < 0.0:
xn = 0.0
if verbose:
print "Estimated variance of phenotypes for specified heriability: ", xn
return xn
if verbose:
print "Using Newton's method to find polynomial roots..."
estimated_variance = newton(p)
new_phenotypes = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
new_phenotypes.append(random.normalvariate(current_mean + each, estimated_variance))
count += 1
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), new_phenotypes)))
f.close()
numpy.savetxt(filename + "_kt_ote2.txt", numpy.column_stack((loci, numpy.array(effects))))
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:], shortopts="vhs:n:l:e:f:i:", longopts=["verbose", "help", "size=",
"number=", "loci=", "effect=",
"filename=", "herit="])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
has_filename = False
has_heritability = False
filename = "my"
heritability = 0.2
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print ("Verbose mode")
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-s", "--size"):
individuals = o[1]
if verbose:
print "Population size is set at", individuals
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
global loci
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
has_filename = True
if verbose:
print "File will be saved as:", filename
elif o[0] in ("-i", "--herit"):
heritability = float(o[1])
has_heritability = True
if verbose:
print "Heritability for simulation specified as:", heritability
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=int(individuals), loci=int(number), infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(), sim.InitGenotype(prop=[0.7, 0.3])], matingScheme=sim.RandomMating(),
postOps=[sim.PyQuanTrait(loci=loci, func=additive_model, infoFields=["qtrait"])],
gen=5)
if verbose:
print "Coalescent process complete. Population evolved."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
#print i.genotype()
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
#print i.qtrait
def fun(sigma, h):
x_exact = phenotypes
x_random = list()
for each in phenotypes:
x_random.append(random.normalvariate(each, sigma))
r = pearsonr(x_exact, x_random)[0]
return r - math.sqrt(h)
#print fun(2.25, 0.25)
if verbose:
print "Building polynomial model for variance tuning..."
points = list()
for i in drange(0, max(effects)*10, 0.001):
points.append(i)
y_points = list()
for i in points:
y_points.append(fun(i, heritability))
z = numpy.polyfit(x=points, y=y_points, deg=3)
p = numpy.poly1d(z)
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations..."
sys.exit()
count += 1
xn = xn - p(xn)/p_d(xn)
if verbose:
print "Estimated variance of phenotypes for specified heriability using Newton's method: ", xn
return xn
if verbose:
print "Using Newton's method to find polynomial roots..."
estimated_variance = newton(p)
new_phenotypes = list()
for each in phenotypes:
new_phenotypes.append(random.normalvariate(each, estimated_variance))
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), new_phenotypes)))
f.close()
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def main():
# First, grab and interpret command line arguments
parser=argparse.ArgumentParser(description="Command line arguments for Simulate")
parser.add_argument("-v", "--verbose", help="Triggers verbose mode", action="store_true")
parser.add_argument("-d", "--distribution", default="normal", choices=["normal", "gamma"], help="Distribution option")
parser.add_argument("-p1", "--parameter1", default=3.0, type=float, help="Shape parameter (only used if distribution choice is gamma)")
parser.add_argument("-p2", "--parameter2", default=1.5, type=float, help="Scale parameter (only used if distribution choice is gamma)")
parser.add_argument("-s", "--size", required=True, type=int, nargs="+", help="Specify population size(s)")
parser.add_argument("-l", "--loci", required=True, type=int, nargs="+", help="Loci with effects")
parser.add_argument("-n", "--number", required=True, default=3000, type=int, help="Number of loci per population or sub-population")
parser.add_argument("-e", "--effect", required=True, type=float, nargs="+", help="Effect size(s) for the loci specified")
parser.add_argument("-i", "--heritability", default=0.2, type=restricted_float, help="Heritability coefficient for population")
parser.add_argument("-m", "--mean", default=2.0, type=float, nargs="+", help="Mean(s) for population phenotype(s)")
parser.add_argument("-g", "--gen", default=5, type=int, help="Number of generations for population to evolve")
parser.add_argument("-r", "--rrate", default=0.0, type=restricted_float, help="Recombination rate for given population")
parser.add_argument("-f", "--filename", default="my", type=str, help="Prefix for output file set")
args = parser.parse_args()
verbose = args.verbose
if verbose:
print "Verbose mode"
distribution = args.distribution
if verbose:
print "Simulation will occur with "+distribution+" distribution"
parameter1 = args.parameter1
if verbose and distribution=="gamma":
print "Gamma distrbution will occur with alpha parameter:", parameter1
parameter2 = args.parameter2
if verbose and distribution=="gamma":
print "Gamma distribution will occur with beta parameter", parameter2
individuals = args.size
if verbose:
print "Population size(s) set at", individuals
loci = args.loci
if verbose:
print "Loci positions per individual set as", loci
number = args.number
if verbose:
print "Number of loci per population set as", number
global effects
effects = args.effect
if verbose:
print "Effects for loci per individual are", effects
heritability = args.heritability
mean = args.mean
if len(mean) == 1 and len(individuals) > 1:
mean = numpy.array(mean)
mean = numpy.repeat(mean, len(individuals), axis=0)
mean = list(mean)
if verbose:
print "Population mean(s) set as", mean
gen = args.gen
if verbose:
print "Number of generations to evolve set as", gen
rrate = args.rrate
if verbose:
print "Recombination rate set as", rrate
filename = args.filename
if verbose:
"File will be saved as", filename
## Start quantitative trait simulation via simuPOP
if verbose:
print "Creating population..."
pop = sim.Population(size=individuals, loci=int(number), infoFields=["qtrait"])
if verbose:
print "Evolving population..."
type(gen)
pop.evolve(initOps=[sim.InitSex(), sim.InitGenotype(prop=[0.7, 0.3])], matingScheme=sim.RandomMating(ops=sim.Recombinator(rates=rrate)),postOps=[sim.PyQuanTrait(loci=loci, func=additive_model, infoFields=["qtrait"])], gen=gen)
if verbose:
print "Coalescent process complete. Population evolved with", pop.numSubPop(),"sub-populations."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
# fun() obtains the heritability equation set to zero for various settings of sigma (standard deviation)
def fun(sigma, h):
x_exact = list()
count = 0
for i in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_exact.append(current_mean + i)
count += 1
x_random = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_random.append(random.normalvariate(current_mean + each, sigma))
count += 1
r = pearsonr(x_exact, x_random)[0]
return r - math.sqrt(h)
if verbose:
print "Building polynomial model for variance tuning..."
# Polyfit fits a polynomial model in numpy to the values obtained from the fun() function
points = list()
for i in drange(0, max(effects)*10, 0.001):
points.append(i)
y_points = list()
for i in points:
y_points.append(fun(i, heritability))
z = numpy.polyfit(x=points, y=y_points, deg=3)
p = numpy.poly1d(z)
# Netwon's method finds the polynomial model's roots
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations...\nPlease choose different settings."
sys.exit()
count += 1
xn = xn - p(xn)/p_d(xn)
if xn < 0.0:
xn = 0.0
if verbose:
print "Estimated variance of phenotypes for specified heriability: ", xn
return xn
if verbose:
print "Using Newton's method to find polynomial roots..."
# Files are saved to the specified location
estimated_variance = newton(p)
new_phenotypes = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
if distribution=="normal":
new_phenotypes.append(random.normalvariate(current_mean + each, estimated_variance))
elif distribution=="gamma":
new_phenotypes.append(random.gammavariate((current_mean + each)/parameter2, numpy.sqrt(estimated_variance/parameter1)))
count += 1
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), new_phenotypes)))
f.close()
numpy.savetxt(filename + "_kt_ote.txt", numpy.column_stack((loci, numpy.array(effects))), fmt='%i %10.7f')
saveCSV(pop, filename + "_genomes.csv")
# Call the convert.R script to convert the output into usable PLINK files
# Will probably need to change this line to something more generalizable in the near future
os.system("Rscript convert.R "+filename)
print "\n\n"
def main():
# First, grab and interpret command line arguments
parser = argparse.ArgumentParser(
description="Command line arguments for Simulate")
parser.add_argument("-v",
"--verbose",
help="Triggers verbose mode",
action="store_true")
parser.add_argument("-d",
"--distribution",
default="normal",
choices=["normal", "gamma"],
help="Distribution option")
parser.add_argument(
"-p1",
"--parameter1",
default=3.0,
type=float,
help="Shape parameter (only used if distribution choice is gamma)")
parser.add_argument(
"-p2",
"--parameter2",
default=1.5,
type=float,
help="Scale parameter (only used if distribution choice is gamma)")
parser.add_argument("-s",
"--size",
required=True,
type=int,
nargs="+",
help="Specify population size(s)")
parser.add_argument("-l",
"--loci",
required=True,
type=int,
nargs="+",
help="Loci with effects")
parser.add_argument("-n",
"--number",
required=True,
default=3000,
type=int,
help="Number of loci per population or sub-population")
parser.add_argument("-e",
"--effect",
required=True,
type=float,
nargs="+",
help="Effect size(s) for the loci specified")
parser.add_argument("-i",
"--heritability",
default=0.2,
type=restricted_float,
help="Heritability coefficient for population")
parser.add_argument("-m",
"--mean",
default=2.0,
type=float,
nargs="+",
help="Mean(s) for population phenotype(s)")
parser.add_argument("-g",
"--gen",
default=5,
type=int,
help="Number of generations for population to evolve")
parser.add_argument("-r",
"--rrate",
default=0.0,
type=restricted_float,
help="Recombination rate for given population")
parser.add_argument("-f",
"--filename",
default="my",
type=str,
help="Prefix for output file set")
args = parser.parse_args()
verbose = args.verbose
if verbose:
print "Verbose mode"
distribution = args.distribution
if verbose:
print "Simulation will occur with " + distribution + " distribution"
parameter1 = args.parameter1
if verbose and distribution == "gamma":
print "Gamma distrbution will occur with alpha parameter:", parameter1
parameter2 = args.parameter2
if verbose and distribution == "gamma":
print "Gamma distribution will occur with beta parameter", parameter2
individuals = args.size
if verbose:
print "Population size(s) set at", individuals
loci = args.loci
if verbose:
print "Loci positions per individual set as", loci
number = args.number
if verbose:
print "Number of loci per population set as", number
global effects
effects = args.effect
if verbose:
print "Effects for loci per individual are", effects
heritability = args.heritability
mean = args.mean
if len(mean) == 1 and len(individuals) > 1:
mean = numpy.array(mean)
mean = numpy.repeat(mean, len(individuals), axis=0)
mean = list(mean)
if verbose:
print "Population mean(s) set as", mean
gen = args.gen
if verbose:
print "Number of generations to evolve set as", gen
rrate = args.rrate
if verbose:
print "Recombination rate set as", rrate
filename = args.filename
if verbose:
"File will be saved as", filename
## Start quantitative trait simulation via simuPOP
if verbose:
print "Creating population..."
pop = sim.Population(size=individuals,
loci=int(number),
infoFields=["qtrait"])
if verbose:
print "Evolving population..."
type(gen)
pop.evolve(
initOps=[sim.InitSex(),
sim.InitGenotype(prop=[0.7, 0.3])],
matingScheme=sim.RandomMating(ops=sim.Recombinator(rates=rrate)),
postOps=[
sim.PyQuanTrait(loci=loci,
func=additive_model,
infoFields=["qtrait"])
],
gen=gen)
if verbose:
print "Coalescent process complete. Population evolved with", pop.numSubPop(
), "sub-populations."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
phenotypes = list()
for i in pop.individuals():
phenotypes.append(i.qtrait)
# fun() obtains the heritability equation set to zero for various settings of sigma (standard deviation)
def fun(sigma, h):
x_exact = list()
count = 0
for i in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_exact.append(current_mean + i)
count += 1
x_random = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
x_random.append(random.normalvariate(current_mean + each, sigma))
count += 1
r = pearsonr(x_exact, x_random)[0]
return r - math.sqrt(h)
if verbose:
print "Building polynomial model for variance tuning..."
# Polyfit fits a polynomial model in numpy to the values obtained from the fun() function
points = list()
for i in drange(0, max(effects) * 10, 0.001):
points.append(i)
y_points = list()
for i in points:
y_points.append(fun(i, heritability))
z = numpy.polyfit(x=points, y=y_points, deg=3)
p = numpy.poly1d(z)
# Netwon's method finds the polynomial model's roots
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations...\nPlease choose different settings."
sys.exit()
count += 1
xn = xn - p(xn) / p_d(xn)
if xn < 0.0:
xn = 0.0
if verbose:
print "Estimated variance of phenotypes for specified heriability: ", xn
return xn
if verbose:
print "Using Newton's method to find polynomial roots..."
# Files are saved to the specified location
estimated_variance = newton(p)
new_phenotypes = list()
count = 0
for each in phenotypes:
current_mean = mean[pop.subPopIndPair(count)[0]]
if distribution == "normal":
new_phenotypes.append(
random.normalvariate(current_mean + each, estimated_variance))
elif distribution == "gamma":
new_phenotypes.append(
random.gammavariate(
(current_mean + each) / parameter2,
numpy.sqrt(estimated_variance / parameter1)))
count += 1
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), new_phenotypes)))
f.close()
numpy.savetxt(filename + "_kt_ote.txt",
numpy.column_stack((loci, numpy.array(effects))),
fmt='%i %10.7f')
saveCSV(pop, filename + "_genomes.csv")
# Call the convert.R script to convert the output into usable PLINK files
# Will probably need to change this line to something more generalizable in the near future
os.system("Rscript convert.R " + filename)
print "\n\n"
def main():
## Check for arguments passed
try:
opts, args = getopt.getopt(sys.argv[1:], shortopts="vhs:n:l:e:f:i:", longopts=["verbose", "help", "size=",
"number=", "loci=", "effect=",
"filename=", "herit="])
except getopt.GetoptError as err:
print(err)
usage()
sys.exit()
verbose = False
has_filename = False
print "\n"
for o in opts:
if o[0] in ("-v", "--verbose"):
verbose = True
print ("Verbose mode")
for o in opts:
if o[0] in ("-h", "--help"):
usage()
sys.exit()
elif o[0] in ("-s", "--size"):
individuals = o[1]
if verbose:
print "Population size is set at", individuals
elif o[0] in ("-n", "--number"):
number = o[1]
if verbose:
print "Number of loci per individual is set at", number
elif o[0] in ("-l", "--loci"):
global loci
loci = o[1].split(",")
loci = map(int, loci)
if verbose:
print "Loci positions per individual are:", loci
elif o[0] in ("-e", "--effect"):
global effects
effects = o[1].split(",")
effects = map(float, effects)
if verbose:
print "Effects for loci per individual are:", effects
elif o[0] in ("-f", "--filename"):
filename = o[1]
has_filename = True
if verbose:
print "File will be saved as:", filename
elif o[0] in ("-i", "--herit"):
heritability = float(o[1])
has_heritability = True
if verbose:
print "Heritability for simulation specified as:", heritability
## Start quantitative trait simulation
if verbose:
print "Creating population..."
pop = sim.Population(size=int(individuals), loci=int(number), infoFields=["qtrait"])
if verbose:
print "Evolving population..."
pop.evolve(initOps=[sim.InitSex(), sim.InitGenotype(prop=[0.7, 0.3])], matingScheme=sim.RandomMating(),
#postOps=[sim.PyQuanTrait(loci=loci, func=trait, infoFields=["qtrait"])],
gen=5)
if verbose:
print "Population evolved."
genotypes = list()
for i in pop.individuals():
genotypes.append(i.genotype())
#print i.genotype()
def fun(sigma2, h):
exact_traits = list()
for i in genotypes:
exact_traits.append(exact_trait(i))
prob_traits = list()
for i in genotypes:
prob_traits.append(prob_trait(i, sigma2))
r = pearsonr(exact_traits, prob_traits)[0]
return r - math.sqrt(h)
def newton(p):
xn = 100
p_d = p.deriv()
count = 0
while abs(p(xn)) > 0.01:
if count > 1000:
print "Unable to converge after 1000 iterations..."
sys.exit()
count += 1
xn = xn - p(xn)/p_d(xn)
if verbose:
print "Estimated variance using Newton's method for solving roots is: ", xn
return xn
## Make sure to change to 100 points around the average "effects"
my_points = list()
for i in range(100):
my_points.append(fun(i, heritability))
z = numpy.polyfit(x=my_points, y=range(100), deg=2)
#print z
p = numpy.poly1d(z)
if verbose:
print "Polynomial fit for finding tuning parameter to match heritability: \n", p
estimated_variance = newton(p)
#print estimated_variance
phenotypes = list()
for i in pop.individuals():
phenotypes.append(prob_trait(i.genotype(), estimated_variance))
#print phenotypes
if has_filename is False:
filename = "my"
f = open(filename + "_qtrait.txt", "w")
f.write("\n".join(map(lambda x: str(x), phenotypes)))
f.close()
saveCSV(pop, filename + "_genomes.csv")
print "\n\n"
def qtrait(geno):
trait = random.normalvariate(sum(geno)*5, random.uniform(0.0001, 3))
if trait <= 0:
trait = random.uniform(0.0001, 1)
return trait
pop.evolve(
initOps=[
sim.InitSex(),
sim.InitGenotype(prop=[0.7, 0.3])
],
matingScheme=sim.RandomMating(),
postOps=[sim.PyQuanTrait(loci=(0, 1, 2, 3, 4, 5, 6, 10, 100), func=qtrait, infoFields=["qtrait"])],
gen=10
)
geno = list()
for i in pop.individuals():
geno.append(i.genotype())
pheno = list()
for i in pop.individuals():
pheno.append(i.qtrait)
f = open("qtrait1.txt", "w")
f.write("\n".join(map(lambda x: str(x), pheno)))
f.close()
saveCSV(pop,"sample.csv")
ploidy1 = [
alleleNames.index(fields[x]) for x in range(2, len(fields), 2)
]
ind.setGenotype(ploidy0, 0)
ind.setGenotype(ploidy1, 1)
ind.setSex(sex)
# close the file
data.close()
return pop
from simuPOP.utils import saveCSV
pop = sim.Population(size=[10],
loci=[3, 2],
lociNames=['rs1', 'rs2', 'rs3', 'rs4', 'rs5'],
alleleNames=['A', 'B'])
sim.initSex(pop)
sim.initGenotype(pop, freq=[0.5, 0.5])
# output sex but not affection status.
saveCSV(pop,
filename='sample.csv',
affectionFormatter=None,
sexFormatter={
sim.MALE: 1,
sim.FEMALE: 2
})
# have a look at the file
print(open('sample.csv').read())
pop1 = importData('sample.csv')
sim.dump(pop1)
