def evaluate(tree_file, ref_msa, out_dir):
util.expect_file_exists(tree_file)
util.expect_file_exists(ref_msa)
util.make_path(out_dir)
prefix = "eval"
cmd = []
cmd.append(common.raxml)
cmd.append('--evaluate')
cmd.append('--msa')
cmd.append(ref_msa)
cmd.append('--model')
cmd.append(common.subst_model)
cmd.append('--tree')
cmd.append(tree_file)
cmd.append('--prefix')
cmd.append(prefix)
cmd.append('--threads')
cmd.append(str(common.iqtree_threads))
cmd.append('--blopt')
cmd.append('nr_safe')
cmd.append('--redo')
cmd.append('--force')
cmd.append('perf_threads')
cmd.append("--blmin")
cmd.append(common.raxml_min_bl)
sub.check_call(cmd, cwd=out_dir, stdout=sub.DEVNULL)
modelfile = os.path.join(out_dir, prefix + ".raxml.bestModel")
return modelfile
def convert(in_type, out_type, in_file, out_file): util.expect_file_exists( in_file ) util.clean_file(out_file) cmd = [] cmd.append(common.genesis_convert) cmd.append(in_type) cmd.append(out_type) cmd.append(in_file) cmd.append(out_file) subprocess.check_call(cmd)
def remove_duplicates(input_msa, output_msa, output_json):
util.expect_file_exists(input_msa)
util.clean_file(output_msa)
util.clean_file(output_json)
cmd = []
cmd.append(common.genesis_reduce_duplicates)
cmd.append(input_msa)
cmd.append(output_msa)
cmd.append(output_json)
subprocess.check_call(cmd)
def launch_raxml(alignment,
model,
output_dir,
seed,
starting_trees=1,
parsimony=False,
bs_trees=0,
cores=16,
debug=False,
parse=False):
util.expect_file_exists(alignment)
util.make_path(output_dir)
prefix = os.path.join(output_dir, "raxml")
prefix += "_" + model.replace("+", "")
cmd = []
cmd.append("mpiexec")
cmd.append("-np")
cmd.append(str(cores))
cmd.append(common.raxml)
cmd.append("--msa")
cmd.append(alignment)
cmd.append("--model")
cmd.append(model)
cmd.append("--seed")
cmd.append(str(seed))
cmd.append("--blmin")
cmd.append(common.raxml_min_bl)
if (parse):
cmd.append("--parse")
cmd.append("--tree")
if (starting_trees > 0):
if (not parsimony):
cmd.append("rand{" + str(starting_trees) + "}")
prefix += "_rand" + str(starting_trees)
else:
cmd.append("pars{" + str(starting_trees) + "}")
prefix += "_pars" + str(starting_trees)
if (bs_trees > 0):
cmd.append("--boostrap")
cmd.append("--bs-trees")
cmd.append(str(bs_trees))
prefix += "_bs" + str(bs_trees)
prefix += "_seed" + str(seed)
cmd.append("--prefix")
cmd.append(prefix)
cmd.append('--force')
cmd.append('perf_threads')
launcher.submit(prefix, cmd, cores, debug)
def outgroup_check(jplace_files, out_dir):
for f in jplace_files:
util.expect_file_exists( f )
util.make_path( out_dir )
cmd = []
cmd.append(common.genesis_outgroup_check)
for f in jplace_files:
cmd.append(f)
outfile = os.path.join( out_dir, "outgroup_check.txt" )
with open( outfile, "w+" ) as logfile:
sub.check_call(cmd, stdout=logfile)
return outfile
def trim_separate_align(input_fasta, mode, runsdir, version_dir):
util.expect_file_exists(input_fasta)
util.clean_dir(runsdir)
util.mkdirp(runsdir)
cmd = []
cmd.append(common.preanalysis1)
cmd.append(input_fasta)
cmd.append(mode)
cmd.append(common.scripts_dir)
cmd.append(common.mafft)
cmd.append(common.outgroup_spec)
cmd.append(version_dir)
cmd.append(str(common.available_cores))
print(" ".join(cmd))
subprocess.check_call(cmd, cwd=runsdir)
def split_alignment_outgroups(input_msa, outgroup_spec, out_dir): util.expect_file_exists( input_msa ) util.expect_file_exists( outgroup_spec ) util.make_path( out_dir ) new_ref_msa="reference.fasta" query_msa="query.fasta" cmd = [] cmd.append(common.genesis_remove_sequences) cmd.append(input_msa) cmd.append(outgroup_spec) cmd.append(new_ref_msa) cmd.append(query_msa) sub.check_call(cmd, cwd=out_dir) return os.path.join(out_dir, new_ref_msa), os.path.join(out_dir, query_msa)
sys.path.insert(0, 'scripts')
import common
import placement
import convert
import util
import raxml_launcher
paths = common.Paths( sys.argv )
# tree = paths.raxml_best_tree
modelfile = paths.raxml_best_model
epa_out_dir = paths.epa_runs_dir
hmmer_out_dir = paths.hmmer_runs_dir
try:
util.expect_file_exists( paths.raxml_credible_ml_trees )
except Exception as e:
print("ERROR: Must run iqtree_tests stage of pipeline first")
raise e
# ================================================================
# start by ensuring we have the outgroups aligned against the ref
# ================================================================
util.make_path_clean( epa_out_dir )
# if outgroup is included in the alignment, separate the two from the alignment currently seen as final
if paths.dataset_has_outgroups:
ref_msa = paths.alignment
query_msa = paths.outgroups_file
# ref_msa, query_msa = placement.split_alignment_outgroups( paths.alignment, common.outgroup_spec, epa_out_dir )
with open(paths.raxml_credible_ml_trees) as ml_trees_file:
ml_trees = ml_trees_file.readlines()
total = len(ml_trees) - 1
i = 0
for tree in ml_trees:
print(i, " / ", total)
# subdirs per tree
cur_outdir = os.path.join(runs_dir, str(i))
epa_result_subdir = os.path.join(paths.epa_rooting_dir, str(i))
util.expect_dir_exists(epa_result_subdir)
i += 1
util.make_path_clean(cur_outdir)
# get the same treefile as used in the normal epa runs
tree_file = os.path.join(epa_result_subdir, "tree.newick")
util.expect_file_exists(tree_file)
# fetch the previously created model file
cur_modelfile = os.path.join(epa_result_subdir, "eval.raxml.bestModel")
util.expect_file_exists(cur_modelfile)
# place the wuhan seq
placement.launch_epa(tree_file,
cur_modelfile,
ref_msa,
query_msa,
cur_outdir,
thorough=True)
result_files.append(os.path.join(cur_outdir, "epa_result.jplace"))