def parse_lineages(tax_file, fmt='fasta', style=GTDB):
""" returns taxonomy object """
id_map = {}
root = TaxNode('root', None, None)
tree = {'root': root}
logger.debug("Parsing %s", tax_file)
if style == GTDB:
add_lineage_to_tree = add_gtdb_lineage_to_tree
generate_lineages = generate_gtdb_lineages
else:
add_lineage_to_tree = add_phylodb_lineage_to_tree
generate_lineages = generate_phylodb_lineages
# generate taxonomy tree
for acc, lineage in generate_lineages(tax_file):
# create TaxNode
node = add_lineage_to_tree(lineage, tree)
id_map[acc] = node
logger.debug("Adding id numbers to %d nodes", len(tree))
# assign numeric IDs
i = 0
for node in treeGenerator(root):
i += 1
node.id = i
logger.debug("Added %d id numbers", i)
return Taxonomy(id_map, None, None, tax_file, root)
def writeDumpFiles(rootNode, nodeStream, nameStream):
for node in treeGenerator(rootNode):
nid = node.id
nname = node.name
if node==rootNode:
nparent = node.id
else:
nparent = node.parent.id
if node.rank == 'domain':
nrank = 'superkingdom'
elif node.rank in edl.taxon.ranks:
nrank = node.rank
else:
nrank = "no rank"
nodeStream.write("%s\t|\t%s\t|\t%s\t\n" % (nid,nparent,nrank))
nameStream.write("%s\t|\t%s\t|\t\t|\tscientific name\t\n" % (nid,nname))
def buildSilvaTree(taxFile, fastaFile, logger):
"""
Given a text taxonomy file (lineage <tab> id <tab> rank) and a fasta file with full lineages as the description:
Return the root node from a taxonomy of edl.taxon.Node objects and a mapping from fasta record IDs to taxids.
"""
rankMap=parseMapFile(taxFile, keyCol=0, valueCol=2, skipFirst=0)
silvaTaxidMap=parseMapFile(taxFile, keyCol=0, valueCol=1, valueType=int, skipFirst=0)
# create core of tree from taxonomy text file
silvaTree={}
maxTaxid=max(silvaTaxidMap.values())
for (lineage, rank) in rankMap.items():
node=edl.silva.SilvaTaxNode.addToTreeFromString(lineage.strip("; "), silvaTree)
node.rank = rankMapping.get(rank,rank)
node.ncbi_tax_id = silvaTaxidMap[lineage]
if not isinstance(node.ncbi_tax_id,int):
logger.warn("NCBI taxid is not an int: %s (%s)" % (node.ncbi_tax_id, node.name))
logger.info("Built tree of %d taxa with the largest ID of %d" % (len(silvaTree),maxTaxid))
# Add leaves to tree from lineages in fasta file and build mapping
taxmap={}
for (hitid,lineage) in getOrgsFromSilvaFasta(fastaFile):
node = edl.silva.SilvaTaxNode.addToTreeFromString(lineage, silvaTree)
taxmap[hitid]=node
logger.info("Added nodes from fasta file for a total of %d" % (len(silvaTree)))
rootNode=next(iter(silvaTree.values())).getRootNode()
# make sure everything is OK
for node in treeGenerator(rootNode):
if not isinstance(node.id,int):
if "ncbi_tax_id" in dir(node):
node.id = int(node.ncbi_tax_id)
else:
maxTaxid+=1
node.id=maxTaxid
logger.info("Cleaning up taxmap")
# change nodes in taxmap to IDs
for hitid in taxmap:
taxmap[hitid]=taxmap[hitid].id
return (rootNode, taxmap)