match_info.Lineage = line[5]
match_info.Sample = sample_name
match_info.RefDatabase = ref_database
# Relative abundance
rel_abund = line[0].replace("%", "")
match_info.Abundance = float(rel_abund)
# get Lineage info from NCBI
match_info.TaxId = int(line[4])
if match_info.TaxId != 0:
try:
match_info = fNCBItax.lineage_extractor(
match_info.TaxId, match_info)
except ValueError:
print("""
Warning: %i is not a valid taxid and will not be considered in the analysis
""" % (match_info.TaxId))
else:
print("Note: %i percent of the reads were unclassified." %
(match_info.Abundance))
store_lineage_info.append(match_info)
# output as a SQLite3:
# note that each row iin the report file (one per tax level) will be a diffferent
# row in the sqlite3
def populate_w_tax(in_df, ref_database,species_threshold,genus_threshold,
family_threshold,order_threshold,class_threshold,phylum_threshold):
#defaults:
#species_threshold = 98.41 # Yeast - Vu et al 2016
#genus_threshold = 96.31 # Yeast - Vu et al 2016
#family_threshold = 88.51 # Filamentous fungi - Vu et al 2019
#order_threshold = 81.21 # Filamentous fungi - Vu et al 2019
#class_threshold = 80.91 # Filamentous fungi - Vu et al 2019
#phylum, kingdom and superkingdom = 0 # no data, no filtering
# Make sure all taxa columns are strings (doesn't automatically happen if the first one is None)
in_df = in_df.assign(LCA_TaxId="",Superkingdom="",Kingdom="",Phylum="",Class="",Order="",Family="",Genus="",Species="")
# index == the #template (fungal match)
for index, row in in_df.iterrows():
match_info = cTaxInfo.TaxInfo()
# define the tax. rank based on similarity:
if ref_database == "UNITE":
split_match = re.split (r'(\|| )', index)
qiden = row['Query_Identity']
match_info.Lineage = split_match[12]
# if taxid is knwon:
if split_match[4] != 'unk_taxid':
match_info.TaxId = int(split_match[4])
match_info = fNCBItax.lineage_extractor(match_info.TaxId , match_info)
# Warning about unknown taxids:
else:
print ("")
print ("WARNING: based on accession number, no taxonomic information was found in NCBI for %s" %(match_info.Lineage))
print ("This match will not get NCBItax taxonomic ranks")
print ("")
match_info.TaxId = split_match[4] # 'unk_taxid'
elif ref_database == "RefSeq":
split_match = re.split (r'(\|| )', index)
qiden = row['Query_Identity']
match_info.TaxId = int(split_match[4])
species = split_match[6] + " " + split_match[8]
match_info.Lineage = species
# include info from NCBI:
match_info = fNCBItax.lineage_extractor(match_info.TaxId, match_info)
elif ref_database == "nt":
split_match = re.split (r'(\|| )', index)
qiden = row['Query_Identity']
match_info.Lineage = split_match[4] + " " + split_match[6]
#get taxid from accession number
taxid = split_match[0]
if taxid == 'unk_taxid':
# Warning about unknown taxids:
print ("")
print ("WARNING: no NCBI's taxid found for accession %s" %(match_info.Lineage))
print ("This match will not get taxonomic ranks")
print ("")
else:
match_info.TaxId = int(taxid)
match_info = fNCBItax.lineage_extractor(match_info.TaxId, match_info)
# Populate the df with lineage info and the LCA taxid:
in_df.at[index, 'Superkingdom'] = match_info.Superkingdom
# Assign LCA_taxid. Go to Kingdom if possible:
in_df.at[index, 'LCA_TaxId'] = match_info.Superkingdom_TaxId
in_df.at[index, 'LCA_TaxId'] = match_info.Kingdom_TaxId
# if it matches to uncultured or unclassified fungus, use the Fungi LCA itaxid:
if match_info.Kingdom == 'Fungi':
in_df.at[index, 'LCA_TaxId'] = 4751
in_df.at[index, 'Kingdom'] = match_info.Kingdom
# fill in the rest of the table according to similarity threshold:
if qiden >= phylum_threshold:
in_df.at[index, 'Phylum'] = match_info.Phylum
if match_info.Phylum_TaxId != None:
in_df.at[index, 'LCA_TaxId'] = match_info.Phylum_TaxId
if qiden >= class_threshold:
in_df.at[index, 'Class'] = match_info.Class
if match_info.Class_TaxId != None:
in_df.at[index, 'LCA_TaxId'] = match_info.Class_TaxId
if qiden >= order_threshold:
in_df.at[index, 'Order'] = match_info.Order
if match_info.Order_TaxId != None:
in_df.at[index, 'LCA_TaxId'] = match_info.Order_TaxId
if qiden >= family_threshold:
in_df.at[index, 'Family'] = match_info.Family
if match_info.Family_TaxId != None:
in_df.at[index, 'LCA_TaxId'] = match_info.Family_TaxId
if qiden >= genus_threshold:
in_df.at[index, 'Genus'] = match_info.Genus
if match_info.Genus_TaxId != None:
in_df.at[index, 'LCA_TaxId'] = match_info.Genus_TaxId
if qiden >= species_threshold:
in_df.at[index, 'Species'] = match_info.Species
if match_info.Species_TaxId != None:
in_df.at[index, 'LCA_TaxId'] = match_info.Species_TaxId
return in_df
# store info in a list of objects
store_fungi_info = []
###### convert existing to cTaxInfo class abd include NCBI lineage info
for key, row in fungi_csv.iterrows():
fungus = cTaxInfo.TaxInfo()
fungus.Sample = row['Sample']
fungus.Lineage = row['Lineage']
fungus.TaxId = row['taxid']
fungus.Covergae = row['Coverage']
# include tax info from NCBI:
fungus = fNCBItax.lineage_extractor(fungus.TaxId, fungus)
store_fungi_info.append(fungus)
###### Add to SQL
connection = sqlite3.connect(sql_fp)
cursor = connection.cursor()
# Create a table if it does not exist:
# Note that Order is written with two 'O's, as Order is a sql command
query = """CREATE TABLE IF NOT EXISTS Fungi (Sample text, Lineage text,
Coverage float, TaxID integer, Superkingdom text, Superkingdom_TaxId integer,
Kingdom text, Kingdom_TaxId integer, Phylum text, Phylum_TaxId integer,
Class text, Class_TaxId integer, OOrder text, Order_TaxId integer, Family text,
Family_TaxId integer, Genus text, Genus_TaxId integer, Species text, Species_TaxId integer);"""