def runDNAdist(seqout, outbase):
"""
This part runs DNAdist for the seqboot output
"""
print "running DNA distance"
distout = outbase + ".dnadist"
dnadist_cline = FDNADistCommandline(sequence=seqout,
outfile=distout,
method="j")
stdout, stderr = dnadist_cline()
return distout
def distances_from_alignment(self, filename, DNA=True):
"""Check we can make a distance matrix from a given alignment."""
self.assertTrue(os.path.isfile(filename), "Missing %s" % filename)
if DNA:
cline = FDNADistCommandline(exes["fdnadist"],
method="j",
sequence=filename,
outfile="test_file",
auto=True)
else:
cline = FProtDistCommandline(exes["fprotdist"],
method="j",
sequence=filename,
outfile="test_file",
auto=True)
stdout, strerr = cline()
# biopython can't grok distance matrices, so we'll just check it exists
self.assertTrue(os.path.isfile("test_file"))
def compute_pairwise_matrices():
alignments = os.listdir(
"data/alignments"
) # get list of all pairwise alignments previously completed
fline = FDNADistCommandline() # create FDNADist object
fline.method = "f"
fline.stdout = True
processes = []
matrix_id = 1
end_filename = 24 # filename end before ".phy", see below
for alignment in alignments:
fline.sequence = "data/alignments/" + alignment # set sequence parameter
fline.outfile = alignment[:end_filename] + "_matrix" # to get string ISL_XXXXXX_vs_ISL_YYYYYY_matrix
print(str(fline))
p = subprocess.Popen(str(fline) + " -odirectory2 data/matrices ",
shell=True) # launching subprocess
processes.append((matrix_id, p)) # appending to process list
matrix_id += 1
while processes:
for proc in processes:
if proc[1].poll() is not None: # if process is completed
processes.remove(proc) # remove it
print("Done matrix %d" % (proc[0]))
def build_nj_phylip(alignment, outfile, outgroup, work_dir="."):
"""
build neighbor joining tree of DNA seqs with PHYLIP in EMBOSS
PHYLIP manual
http://evolution.genetics.washington.edu/phylip/doc/
"""
phy_file = op.join(work_dir, "work", "aln.phy")
try:
AlignIO.write(alignment, file(phy_file, "w"), "phylip")
except ValueError:
print >>sys.stderr, \
"Repeated seq name, possibly due to truncation. NJ tree not built."
return None
seqboot_out = phy_file.rsplit(".", 1)[0] + ".fseqboot"
seqboot_cl = FSeqBootCommandline(FPHYLIP_BIN("fseqboot"), \
sequence=phy_file, outfile=seqboot_out, \
seqtype="d", reps=100, seed=12345)
stdout, stderr = seqboot_cl()
logging.debug("Resampling alignment: %s" % seqboot_cl)
dnadist_out = phy_file.rsplit(".", 1)[0] + ".fdnadist"
dnadist_cl = FDNADistCommandline(FPHYLIP_BIN("fdnadist"), \
sequence=seqboot_out, outfile=dnadist_out, method="f")
stdout, stderr = dnadist_cl()
logging.debug\
("Calculating distance for bootstrapped alignments: %s" % dnadist_cl)
neighbor_out = phy_file.rsplit(".", 1)[0] + ".njtree"
e = phy_file.rsplit(".", 1)[0] + ".fneighbor"
neighbor_cl = FNeighborCommandline(FPHYLIP_BIN("fneighbor"), \
datafile=dnadist_out, outfile=e, outtreefile=neighbor_out)
stdout, stderr = neighbor_cl()
logging.debug("Building Neighbor Joining tree: %s" % neighbor_cl)
consense_out = phy_file.rsplit(".", 1)[0] + ".consensustree.nodesupport"
e = phy_file.rsplit(".", 1)[0] + ".fconsense"
consense_cl = FConsenseCommandline(FPHYLIP_BIN("fconsense"), \
intreefile=neighbor_out, outfile=e, outtreefile=consense_out)
stdout, stderr = consense_cl()
logging.debug("Building consensus tree: %s" % consense_cl)
# distance without bootstrapping
dnadist_out0 = phy_file.rsplit(".", 1)[0] + ".fdnadist0"
dnadist_cl0 = FDNADistCommandline(FPHYLIP_BIN("fdnadist"), \
sequence=phy_file, outfile=dnadist_out0, method="f")
stdout, stderr = dnadist_cl0()
logging.debug\
("Calculating distance for original alignment: %s" % dnadist_cl0)
# infer branch length on consensus tree
consensustree1 = phy_file.rsplit(".", 1)[0] + ".consensustree.branchlength"
run_ffitch(distfile=dnadist_out0, outtreefile=consensustree1, \
intreefile=consense_out)
# write final tree
ct_s = Tree(consense_out)
if outgroup:
t1 = consensustree1 + ".rooted"
t2 = smart_reroot(consensustree1, outgroup, t1)
if t2 == t1:
outfile = outfile.replace(".unrooted", "")
ct_b = Tree(t2)
else:
ct_b = Tree(consensustree1)
nodesupport = {}
for node in ct_s.traverse("postorder"):
node_children = tuple(sorted([f.name for f in node]))
if len(node_children) > 1:
nodesupport[node_children] = node.dist / 100.
for k, v in nodesupport.items():
ct_b.get_common_ancestor(*k).support = v
print ct_b
ct_b.write(format=0, outfile=outfile)
try:
s = op.getsize(outfile)
except OSError:
s = 0
if s:
logging.debug("NJ tree printed to %s" % outfile)
return outfile, phy_file
else:
logging.debug("Something was wrong. NJ tree was not built.")
return None
infile_Formatted.write("\n")
with open("query_searchSequences_dna.fasta",'r') as infile_sequences:
for record in SeqIO.parse("query_searchSequences_dna.fasta", "fasta"):
speciesName = '{:<10}'.format(record.id[:10]) #This line ensures that any species' name that is greater than 10 characters, gets cut to 10 characters. And any species' name that is less than
#10 characters, gets increased to 10 characters using spaces.
speciesSequence = record.seq[:numberOfPositions] #This line cuts each sequence to the length stored in 'numberOfPositions' so that each sequence is the same length.
infile_Formatted.write(str(speciesName))
infile_Formatted.write(str(speciesSequence))
infile_Formatted.write("\n")
##The next part of the code uses FDNADistCommandline and the "FormattedInputFile.txt" file created above to create a distance matrix.
#The first part is "/usr/local/bin/fdnadist", which tells Python where it can find this application. The sequence parameter tells FDNADistCommandline which file you would like to use that contains
#the sequences you want to compute a distance matrix for. The method parameter tells FDNADistCommandline which distance matrix algorithm you would like to use; in this case, f means FDNADistCommandline
#will use the F84 distance model. The outfile parameter specifies the name of the outfile in which FDNADistCommandline will write the results to.
FDNADist_matrix = FDNADistCommandline("/usr/local/bin/fdnadist",sequence="FormattedInputFile.txt",method="f",outfile="distanceMatrix.fdnadist")
stdout, stderr = FDNADist_matrix() #This line is required in order for FDNADistCommandline to actually write anything to the outfile.
##The next part of the code uses FNeighborCommandline and the "distanceMatrix.fdnadist" file created above to create a phylogenetic tree.
#The first part is "/usr/local/bin/fneighbor", which tells Python where it can find this application. The datafile parameter tells FNeighborCommandline which file you would like to use that contains
#the distance matrix that will be used to create the phylogenetic tree. The outfile parameter specifies the name of the outfile in which FNeighborCommandline will write the results to.
FNeighbor_tree = FNeighborCommandline("/usr/local/bin/fneighbor",datafile="distanceMatrix.fdnadist",outfile="treeFile.fneighbor")
stdout, stderr = FNeighbor_tree() #This line is required in order for FNeighborCommandline to actually write anything to the outfile.
##The next part of the code creates alignments between each and every one of the sequences in the "query_searchSequences_dna.fasta" using the NeedleallCommandline.
#The next few lines create an input file for NeedleallCommandline. The FASTA file called 'query_searchSequences_dna.fasta' cannot be used because it contains spaces between each record and for
#some reason the spaces affect the Needleall alignemnt. So the next few lines create a FASTA without spaces between each record called 'needlallAlignmentInput_Nucleotide.fasta'. The sequences are also all