def optParse():
# This function handles the command line options and does some error checking.
parser = argparse.ArgumentParser(description="Runs codeml on a directory full of .fa files. Files MUST have .fa extension. Dependencies: PAML, newickutils (if you want to prune your tree with --prune)");
parser.add_argument("-i", dest="input", help="Input. A directory containing many FASTA (.fa) files.");
parser.add_argument("-p", dest="paml_path", help="You must specify the full path to your PAML DIRECTORY here.");
parser.add_argument("-t", dest="tree_file", help="A user specified tree for codeml to use. If not specified, codeml will infer the tree.", default="");
parser.add_argument("--prune", dest="prune_opt", help="If not all species present in the tree will be present in each alignment, set this flag to prune the tree for each file.", action="store_true");
# parser.add_argument("-seqtype", dest="paml_seqtype", help="Enter either 'codon' or 'aa'. Default value is 'codon'.", default='codon');
parser.add_argument("-v", dest="verbosity", help="An option to control the output printed to the screen. 1: print all codeml output, 0: print only a progress bar. Default: 1", type=int, default=1);
parser.add_argument("-o", dest="output", help="Desired output directory. If none is entered, will be determined automatically.", default=False);
args = parser.parse_args();
if args.input == None or args.paml_path == None:
sys.exit(gc.errorOut(1, "Both -i must be set."));
if not os.path.isdir(args.input) or not os.path.isdir(args.paml_path):
sys.exit(gc.errorOut(2, "Both -i must be valid directory paths!"));
else:
args.input = os.path.abspath(args.input);
args.paml_path = os.path.abspath(args.paml_path);
if not os.path.isfile(args.tree_file):
sys.exit(gc.errorOut(3, "-t must be a valid file name."));
try:
td, tree, r = gt.treeParse(open(args.tree_file, "r").read().replace("\n",""));
except:
sys.exit(gc.errorOut(4, "-t does not contain a valid Newick string!"));
if args.verbosity not in [0,1]:
sys.exit(gc.errorOut(6, "-v must take values of either 1 or 0"));
return args.input, args.paml_path, args.tree_file, args.prune_opt, args.verbosity, args.output;
def splitThreads(arglist):
filelist_func = arglist[0];
orig_targets = arglist[1];
u = arglist[2];
p = arglist[3];
threads = arglist[4];
results_dict = {};
for filename in filelist_func:
if ".fa" not in filename:
continue;
print filename;
if u != 1:
gid = filename[:filename.index("_ancprobs.fa")];
elif u == 1:
gid = filename[:filename.index(".fa")];
gene = "_".join(gid.split("_")[:2]);
chromosome = gid[gid.find("chr"):gid.find("chr")+4]
infilename = os.path.join(indir, filename);
if u != 1:
treefilename = os.path.join(indir, gid + "_anc.tre");
tree = open(treefilename,"r").read().replace("\n","");
tree_dict, new_tree = gwctree.treeParse(tree);
if orig_targets != "":
results_key = str(orig_targets);
if results_key not in results_dict:
results_dict[results_key] = [[],[],[]];
targets = copy.deepcopy(orig_targets);
#Resets the targets for each gene.
results_dict = convergence.convCheck(infilename, results_dict, results_key, targets, prob_thresh, chromosome, gene, tree_dict, u);
#Checking for convergent sites
else:
target_nodes = getTargs(tree_dict, p);
for targets in target_nodes:
if tree_dict[targets[0]][1] == targets[1] or tree_dict[targets[1]][1] == targets[0]:
continue;
#If one node is the ancestor of the other, skip this comparison.
node_key = "";
for n in targets:
if "_" in n:
node_key = node_key + n[n.index("_")+1:];
else:
node_key = node_key + tree_dict[n][3];
if n == targets[0]:
node_key = node_key + "-";
if node_key not in results_dict:
results_dict[node_key] = [[],[],[]];
results_dict = convergence.convCheck(infilename, results_dict, node_key, targets, prob_thresh, chromosome, gene, tree_dict, u);
return results_dict;
gwctcore.logCheck(l, logfilename, "-------------------------------------");
#sys.exit();
if not os.path.exists(outdir):
gwctcore.logCheck(l, logfilename, gwctcore.getTime() + " | Creating codeml output directory:\t" + outdir);
cmd = "mkdir " + outdir;
os.system(cmd);
if aopt == 1:
if not os.path.exists(ancdir):
gwctcore.logCheck(l, logfilename, gwctcore.getTime() + " | Creating directory to pass ancestral sequences and trees:\t" + ancdir);
cmd = "mkdir " + ancdir;
os.system(cmd);
if prune == 1:
gwctcore.logCheck(l, logfilename, gwctcore.getTime() + " | Retrieving tree info...");
td, tree = gwctree.treeParse(open(treefile, "r").read().replace("\n",""),0);
tips = [];
for node in td:
if td[node][2] == 'tip':
tips.append(node);
gwctcore.logCheck(l, logfilename, gwctcore.getTime() + " | Starting codeml runs...\n");
if v == 0:
codeml_logfile = os.path.join(script_outdir, "codeml.log");
ctlfilename = "codeml.ctl";
i = 0;
numbars = 0;
donepercent = [];
def splitThreads(arglist):
filelist_func = arglist[0]
orig_targets = arglist[1]
p = arglist[2]
threads = arglist[3]
results_dict = {}
for filename in filelist_func:
print filename
if "-ancprobs.fa" in filename:
gid = filename[:filename.index("-ancprobs.fa")]
treefilename = os.path.join(ancdir, gid + "-anc.tre")
else:
gid = filename[:filename.index("_ancprobs.fa")]
treefilename = os.path.join(ancdir, gid + "_anc.tre")
infilename = os.path.join(ancdir, filename)
tree = open(treefilename, "r").read().strip()
tree_dict, new_tree, root = gt.treeParse(tree)
#gene = "-".join(gid.split("_")[:2]);
#chromosome = gid[gid.find("chr"):gid.find("chr")+4]
if orig_targets != "":
results_key = str(orig_targets)
if results_key not in results_dict:
results_dict[results_key] = [[], [], []]
targets = copy.deepcopy(orig_targets)
#Resets the targets for each gene.
results_dict = convergence.convCheck(infilename, results_dict,
results_key, targets,
prob_thresh, gid, tree_dict,
pairwise)
#Checking for convergent sites
else:
target_nodes = getTargs(tree_dict, p)
for targets in target_nodes:
if tree_dict[targets[0]][1] == targets[1] or tree_dict[
targets[1]][1] == targets[0]:
continue
# If one node is the ancestor of the other, skip this comparison.
node_key = ""
for n in targets:
if "_" in n:
node_key += n[n.index("_") + 1:]
else:
node_key += tree_dict[n][3]
if n == targets[0]:
node_key += "-"
if node_key not in results_dict:
results_dict[node_key] = [[], [], []]
print targets
targets = [[t] for t in targets]
print targets
results_dict = convergence.convCheck(infilename, results_dict,
node_key, targets,
prob_thresh, gid,
tree_dict, pairwise)
return results_dict
gc.printWrite(logfilename, " -> Printing all codeml output to the screen (-v 1)");
else:
gc.printWrite(logfilename, " -> Silent mode. Not printing codeml output to the screen (-v 0)");
gc.printWrite(logfilename, "-------------------------------------");
# Print IO info to screen for user.
filelist = os.listdir(indir);
print "+ Creating codeml output directory:\t" + codemldir;
os.system("mkdir " + codemldir);
print "+ Creating directory to pass ancestral sequences and trees:\t" + ancdir;
os.system("mkdir " + ancdir);
# Create output directories.
if prune:
print " -> Retrieving tree info...";
td, tree, r = gt.treeParse(open(treefile, "r").read().replace("\n",""),0);
tips = [node for node in td if td[node][2] == 'tip'];
# Read tree info for pruning
gc.printWrite(logfilename, gc.getTime() + " | Starting codeml runs...\n");
if v == 0:
codeml_logfile = os.path.join(outdir, "codeml.stdout");
ctlfilename = "codeml.ctl";
i, numbars, donepercent, numfiles = 0, 0, [], len(filelist);
# Loading bar stuff
fa_skip = [];
for cur_file in filelist:
if v == 0:
numbars, donepercent = gc.loadingBar(i, numfiles, donepercent, numbars);
def splitThreads(arglist):
filelist_func = arglist[0]
orig_targets = arglist[1]
u = arglist[2]
p = arglist[3]
threads = arglist[4]
results_dict = {}
for filename in filelist_func:
if ".fa" not in filename:
continue
print filename
if u != 1:
gid = filename[:filename.index("_ancprobs.fa")]
elif u == 1:
gid = filename[:filename.index(".fa")]
gene = "_".join(gid.split("_")[:2])
chromosome = gid[gid.find("chr"):gid.find("chr") + 4]
infilename = os.path.join(indir, filename)
if u != 1:
treefilename = os.path.join(indir, gid + "_anc.tre")
tree = open(treefilename, "r").read().replace("\n", "")
tree_dict, new_tree = gwctree.treeParse(tree)
if orig_targets != "":
results_key = str(orig_targets)
if results_key not in results_dict:
results_dict[results_key] = [[], [], []]
targets = copy.deepcopy(orig_targets)
#Resets the targets for each gene.
results_dict = convergence.convCheck(infilename, results_dict,
results_key, targets,
prob_thresh, chromosome, gene,
tree_dict, u)
#Checking for convergent sites
else:
target_nodes = getTargs(tree_dict, p)
for targets in target_nodes:
if tree_dict[targets[0]][1] == targets[1] or tree_dict[
targets[1]][1] == targets[0]:
continue
#If one node is the ancestor of the other, skip this comparison.
node_key = ""
for n in targets:
if "_" in n:
node_key = node_key + n[n.index("_") + 1:]
else:
node_key = node_key + tree_dict[n][3]
if n == targets[0]:
node_key = node_key + "-"
if node_key not in results_dict:
results_dict[node_key] = [[], [], []]
results_dict = convergence.convCheck(infilename, results_dict,
node_key, targets,
prob_thresh, chromosome,
gene, tree_dict, u)
return results_dict
