def getCallInfoData(self, call_method_id, phenotype_method_id):
"""
2009-5-17
add region & country in the returning data structure
2009-2-2
get ecotype id, name, latitude, longitude, pc1, pc2, phenotype for all call_infos in one call_method_id
"""
#1st get call_info_id 2 PC12 mapping
rows = model.Stock_250kDB.CallInfo.query.filter_by(method_id=call_method_id).all()
call_info_id2pcs = {}
for row in rows:
call_info_id2pcs[row.id] = row.pc_value_ls[:2]
#2nd check if model.pheno_data exists
pheno_data = getattr(model, 'pheno_data', None)
if pheno_data is None:
from variation.src.OutputPhenotype import OutputPhenotype
pheno_data = OutputPhenotype.getPhenotypeData(model.db.metadata.bind, phenotype_avg_table=model.Stock_250kDB.PhenotypeAvg.table.name,\
phenotype_method_table=model.Stock_250kDB.PhenotypeMethod.table.name)
model.pheno_data = pheno_data
#3rd finally construct the full data and turn it into json
column_name_type_ls = [("date", ("date", "Date")), ("ecotypeid", ("number", "Ecotype ID")), ("label", ("string","ID Name Phenotype")), \
("lat",("number", "Latitude")), ("lon",("number", "Longitude")), ("name", ("string", "Native Name")), \
("pc1", ("number","PC1")), ("pc2", ("number", "PC2")), ("phenotype", ("number", "Phenotype")),\
("region", ("string", "Region")), ("country", ("string", "Country"))]
description = dict(column_name_type_ls)
rows = model.db.metadata.bind.execute("select * from view_call where call_method_id=%s"%(call_method_id))
return_ls = []
for row in rows:
pc12 = call_info_id2pcs.get(row.call_info_id)
if pc12:
call_info_pc1, call_info_pc2 = pc12[:2]
pc1 = call_info_pc1.pc_value
pc2 = call_info_pc2.pc_value
else:
pc1 = pc2 = 0
pheno_row_index = pheno_data.row_id2row_index.get(row.ecotype_id)
pheno_col_index = pheno_data.col_id2col_index.get(int(phenotype_method_id))
if pheno_row_index is not None and pheno_col_index is not None:
phenotype_value = pheno_data.data_matrix[pheno_row_index][pheno_col_index]
if phenotype_value == 'NA':
phenotype_value = None
else:
phenotype_value = None #numpy.nan can't be recognized by ToJSon()
label = '%s ID:%s Phenotype:%s.'%(row.nativename, row.ecotype_id, phenotype_value)
#label = 'ID:%s. Name=%s. Phenotype=%s.'%(row.ecotype_id, row.nativename, phenotype_value)
return_ls.append(dict(ecotypeid=row.ecotype_id, name=row.nativename, lat=row.latitude, lon=row.longitude,\
pc1=pc1, pc2=pc2, phenotype=phenotype_value, label=label, date=datetime.date(2009,2,3), \
region=row.region, country=row.country))
data_table = gviz_api.DataTable(description)
data_table.LoadData(return_ls)
column_ls = [row[0] for row in column_name_type_ls]
#column_ls.sort() #['date', 'ecotypeid', 'label', 'lat', 'lon', 'name', 'pc1', 'pc2', 'phenotype']
json_result = data_table.ToJSon(columns_order=column_ls) #ToJSonResponse only works with google.visualization.Query
return json_result
def getPhenotypeData(cls):
"""
2009-4-30
get data from all the phenotypes into one matrix (accession by phenotype)
"""
phenoData = getattr(model, "phenoData", None)
if phenoData is None:
from variation.src.OutputPhenotype import OutputPhenotype
phenoData = OutputPhenotype.getPhenotypeData(
model.db.metadata.bind,
phenotype_avg_table=model.Stock_250kDB.PhenotypeAvg.table.name,
phenotype_method_table=model.Stock_250kDB.PhenotypeMethod.table.name,
)
model.phenoData = phenoData
return phenoData
def getPhenotypeValue(self, id=None, returnJson=True):
"""
2009-4-6
given a phenotype method id
return
1. ecotype id : phenotype value
2. min & max phenotype value for requests to make phenotype markers on the map
"""
phenotype_method_id = request.params.get('phenotype_method_id', id)
#2nd check if model.phenoData exists
phenoData = getattr(model, 'phenoData', None)
if phenoData is None:
from variation.src.OutputPhenotype import OutputPhenotype
phenoData = OutputPhenotype.getPhenotypeData(model.db.metadata.bind, phenotype_avg_table=model.Stock_250kDB.PhenotypeAvg.table.name,\
phenotype_method_table=model.Stock_250kDB.PhenotypeMethod.table.name)
model.phenoData = phenoData
ecotype_id2phenotype_value = {}
pheno_col_index = phenoData.col_id2col_index.get(int(phenotype_method_id))
min_value = None
max_value = None
for ecotype_id,pheno_row_index in phenoData.row_id2row_index.iteritems():
if pheno_row_index is not None and pheno_col_index is not None:
phenotype_value = phenoData.data_matrix[pheno_row_index][pheno_col_index]
ecotype_id2phenotype_value[ecotype_id] = phenotype_value
if phenotype_value != 'NA':
if min_value == None or phenotype_value<min_value:
min_value = phenotype_value
if max_value == None or phenotype_value>max_value:
max_value = phenotype_value
dc = dict(min_value=min_value, max_value=max_value, ecotype_id2phenotype_value=ecotype_id2phenotype_value)
if returnJson:
response.headers['Content-Type'] = 'application/json'
return simplejson.dumps(dc)
else:
return dc
def getEcotypeAllelePhenotype(self, id=None):
"""
2009-6-22
if phenotype_method.data_type is binary or ordered_categorical, add 1 to phenotype_value
to make accessions with zero-phenotype-value visible in the "Ecotype Allele Phenotype BarChart"
2009-3-5
return json data of ecotype, allele, phenotype for the javascript in templates/snp.html
"""
self.readInRequestParams(request, c)
#1st get call_info_id 2 PC12 mapping
rows = model.Stock_250kDB.CallInfo.query.filter_by(method_id=c.call_method_id).all()
call_info_id2pcs = {}
for row in rows:
call_info_id2pcs[row.id] = row.pc_value_ls[:10]
#if h.ecotype_info is None: #2009-3-6 not used right now
# h.ecotype_info = getEcotypeInfo(model.db)
pheno_data = getattr(model, 'pheno_data', None)
if pheno_data is None:
from variation.src.OutputPhenotype import OutputPhenotype
pheno_data = OutputPhenotype.getPhenotypeData(model.db.metadata.bind, phenotype_avg_table=model.Stock_250kDB.PhenotypeAvg.table.name,\
phenotype_method_table=model.Stock_250kDB.PhenotypeMethod.table.name)
model.pheno_data = pheno_data
snpData = hc.getSNPDataGivenCallMethodID(c.call_method_id)
pm = PhenotypeMethod.get(c.phenotype_method_id)
column_name_type_ls = [("label", ("string","ID Name Phenotype")), ("date", ("date", "Date")), \
("lon",("number", "Longitude")), ("lat",("number", "Latitude")), \
("ecotypeid", ("number", "Ecotype ID")), ("name", ("string", "Native Name")), \
("phenotype", ("number", "Phenotype")), \
("allele", ("string", "Allele")), ("country", ("string", "Country")),\
("pc1", ("number","PC1")), ("pc2", ("number", "PC2")), ("pc3", ("number","PC3")), \
("pc4", ("number", "PC4")), ("pc5", ("number","PC5")), ("pc6", ("number", "PC6"))]
description = dict(column_name_type_ls)
rows = model.db.metadata.bind.execute("select * from view_call where call_method_id=%s"%(c.call_method_id))
return_ls = []
for row in rows:
pcs = call_info_id2pcs.get(row.call_info_id)
if pcs:
pc_value_ls = [getattr(call_info_pc, 'pc_value') for call_info_pc in pcs]
else:
pc_value_ls = [0]*10
pheno_row_index = pheno_data.row_id2row_index.get(row.ecotype_id)
pheno_col_index = pheno_data.col_id2col_index.get(c.phenotype_method_id)
if pheno_row_index is not None and pheno_col_index is not None:
phenotype_value = pheno_data.data_matrix[pheno_row_index][pheno_col_index]
if phenotype_value == 'NA':
phenotype_value = None
else:
phenotype_value = None #numpy.nan can't be recognized by ToJSon()
if (pm.data_type=='binary' or pm.data_type=='ordered_categorical') and phenotype_value is not None:
phenotype_value += 1
label = '%s ID:%s Phenotype:%s.'%(row.nativename, row.ecotype_id, phenotype_value)
snpdata_row_index = snpData.row_id2row_index.get(str(row.ecotype_id))
snpdata_col_index = snpData.col_id2col_index.get('%s_%s'%(c.chromosome, c.position))
if snpdata_row_index is None or snpdata_col_index is None:
allele = -2
else:
allele = snpData.data_matrix[snpdata_row_index][snpdata_col_index]
allele = number2nt[allele]
return_ls.append(dict(date=datetime.date(2009,2,3), ecotypeid=row.ecotype_id, label=label, name=row.nativename, \
lat=row.latitude, lon=row.longitude,\
pc1=pc_value_ls[0], pc2=pc_value_ls[1], pc3=pc_value_ls[2], pc4=pc_value_ls[3], \
pc5=pc_value_ls[4], pc6=pc_value_ls[5], \
phenotype=phenotype_value, allele=allele, country=row.country))
data_table = gviz_api.DataTable(description)
data_table.LoadData(return_ls)
column_ls = [row[0] for row in column_name_type_ls]
#column_ls.sort() #['date', 'ecotypeid', 'label', 'lat', 'lon', 'name', 'pc1', 'pc2', 'phenotype']
json_result = data_table.ToJSon(columns_order=column_ls) #ToJSonResponse only works with google.visualization.Query
return json_result