#! /usr/bin/env python

import collections
import os, sys, time, argparse
import pandas
import numpy as np
from Bio import SeqIO
from Bio.Seq import Seq
from sklearn.cluster import AffinityPropagation
from Bio.SeqUtils import GC
from Bio.Alphabet import IUPAC
from argparse import RawTextHelpFormatter
import csv
##########################Get GC Count#######################################
def GC_Fasta_file(b,name):
    """Makes a tab-delimeted output indicating the GC count associated with each contig"""
    GC_dict = {}
    input_file = open(os.path.join(b,name), 'r') 
    output_file = open('GC_count','w')
    output_file.write("%s\t%s\n" % ('contig','GC'))
    for cur_record in SeqIO.parse(input_file, "fasta") :
        gene_name = cur_record.id 
        GC_percent = float(GC(cur_record.seq))
        GC_dict.setdefault(gene_name,[])
        GC_dict[gene_name].append(GC_percent) 
        output_line = '%s\t%i\n' % \
        (gene_name, GC_percent) 
        output_file.write(output_line)

        
    output_file.close() 
    input_file.close()
###########################Get tetramer frequencies#####################################

def kmer_list(dna, k):
    """Makes list of k-mers based on input of k and the dna sequence"""
    result = []
    dna = dna.upper()
    dna_edit = Seq(dna,IUPAC.unambiguous_dna)
    reverse_complement = (dna_edit).reverse_complement()
    dna_edit = str(dna_edit)
    reverse_complement = str(reverse_complement)	
    for x in range(len(dna_edit)+1-k):
        result.append(dna_edit[x:x+k])
    for x in range(len(reverse_complement)+1-k):
        result.append(reverse_complement[x:x+k])
    result= [s for s in result if not s.strip('AGTC')]
    return result


def kmer_counts(kmer,b,name):
    kmer_dict = {}
    for record in SeqIO.parse(os.path.join(b,name), "fasta"):
        id_=str(record.id)
        seq = str(record.seq)
        length = len(seq)
        tetra = kmer_list(seq,kmer)           
        c = collections.Counter(tetra)
        c = dict(c)
        val = list(c.values())
        val_edit = []
        for freq in val:
            freq= (float(freq)/float(length))
            val_edit.append(freq)
    
        keys = []
        for key in c:
            keys.append(str(key))
        c_edit = dict(zip(keys,val_edit))
        kmer_dict.setdefault(str(id_),[])
        kmer_dict[str(id_)].append(c_edit)
    return kmer_dict
    
def output(a):
    """Builds tab-delimeted file based on dictionary of tetramers"""
    topleft = 'contig' 
    headers = sorted(set(key
                        for row in a.values()
                        for key in row[0]))
    writer = csv.writer(open('tetramer-frequencies','wb'), delimiter='\t')
    writer.writerow([topleft] + headers)
    for key in a:
        row = [key]
        for header in headers:
            row.append(a[key][0].get(header, 0))
        writer.writerow(row)
#########################Combine tetramer frequencies and GC counts in tab delimeted format and normalize################################
def get_contigs(c):
    GC = open("GC_count","r")
    dataframe =pandas.read_csv(GC,sep="\t")
    dataframe=dataframe.set_index("contig")
    dataframe *= 100
    dataframe +=1  
    dataframe = np.log10(dataframe)
    tetra = open("tetramer-frequencies","r")
    dataframe2=pandas.read_csv(tetra,sep="\t")
    dataframe2=dataframe2.set_index("contig")
    dataframe2*=100
    dataframe2 += 1
    dataframe2 = np.log10(dataframe2)
    cov = open(str(c),"r")
    reader =list(csv.reader(cov,delimiter='\t'))
    titles = ["contig"]
    reader.insert(0,titles)
    dataframe3=pandas.DataFrame(reader)
    dataframe3.columns = dataframe3.iloc[0]
    dataframe3 = dataframe3[1:]
    dataframe3=dataframe3.set_index("contig")
    result = pandas.concat([dataframe, dataframe2,dataframe3],axis=1,join='inner')
    result.to_csv(path_or_buf=str("tetra-GC-out"),header=None,sep ="\t")


##########################Run AP for refinement##############################    
def create_fasta_dict(fastafile):
    """makes dictionary using fasta files to be binned"""
    fasta_dict = {}
    for record in SeqIO.parse(fastafile, "fasta"):
        fasta_dict[record.id] = record.seq
    return fasta_dict


def get_cov_data(h):
    """Makes a dictionary of coverage values for affinity propagation"""
    all_cov_data = {}
    #for line in open(str(sys.argv[1]), "r"):
    for line in open(str(h), "r"):
        line = line.rstrip()
        cov_data = line.split()
        all_cov_data[cov_data[0]] = cov_data
    return all_cov_data


def cov_array(a,b,name,size):
    """Computes a coverage array based on the contigs in files and the contig names associated with the coverage file"""
    names = []
    c = 0
    cov_array = []
    for record in SeqIO.parse(os.path.join(b,name), "fasta"):
        if len(record.seq) >= int(size):
            if record.id in a.keys():
                data = a[record.id]
                names.append(data[0])
                data.remove(data[0])
                line = " ".join(data)
                if c == 1:
                    temparray = np.fromstring(line, dtype=float, sep=' ')
                    cov_array = np.vstack((cov_array, temparray))
                if c == 0:
                    cov_array = np.fromstring(line, dtype=float, sep=' ')
                    c += 1
    print cov_array.shape
    return cov_array, names    

def affinity_propagation(array,names,file_name,damping,iterations,convergence,preference,path,output_directory):
    """Uses affinity propagation to make putative bins"""    
    if os.path.isdir(str(output_directory)) is False:
        os.mkdir(output_directory)
    apclust = AffinityPropagation(damping=float(damping), max_iter=int(iterations), convergence_iter=int(convergence), copy=True, preference=int(preference), affinity='euclidean', verbose=False).fit_predict(array)
    print"""-------------------------------------------------------
                            --Creating Bins--
            -------------------------------------------------------"""
    outfile_data = {}
    i = 0
    while i < len(names):
        if apclust[i] in outfile_data.keys():
            outfile_data[apclust[i]].append(names[i])
        if apclust[i] not in outfile_data.keys():
            outfile_data[apclust[i]] = [names[i]]
        i += 1
    out_name = file_name.rsplit(".",1)[0]
    with open(os.path.join(path,file_name),"r") as input2_file: 
        fasta_dict = create_fasta_dict(input2_file)                
        count = 0                
        for k in outfile_data:
            if len(outfile_data[k]) >= 5:
                output_file = open(os.path.join(output_directory,str(out_name)+"-refined_%s.fna" % (k)), "w" )
                for x in outfile_data[k]:
                    output_file.write(">"+str(x)+"\n"+str(fasta_dict[x])+"\n")
                output_file.close()
                count = count + 1
            elif len(outfile_data[k]) < 5:
                if any((len(fasta_dict[x])>50000) for x in outfile_data[k]):
                    output_file = open(os.path.join(output_directory,str(out_name)+"-refined_%s.fna" % (k)), "w" )
                    for x in outfile_data[k]:
                        output_file.write(">"+str(x)+"\n"+str(fasta_dict[x])+"\n")
                    output_file.close()
                    count = count +1
            print "Cluster "+str(k)+": "+str(len(outfile_data[k]))
        print ("""Total Number of Bins: %i""" % count)
                
class Logger(object):
    def __init__(self):
        self.terminal = sys.stdout
        self.log = open("cluster-log.txt", "a")

    def write(self, message):
        self.terminal.write(message)
        self.log.write(message)  

    def flush(self):
        pass  

    
if __name__ == '__main__':
    parser = argparse.ArgumentParser(prog='Binsanity-refine', description="""
    Binsanity clusters contigs based on coverage and refines these 'bins' using tetramer frequencies and GC content. 
    ----------------------------------------------------------------------------------------------------------------
    Binsanity-refine uses combined coverage and composition 
    (in the form of tetramer frequencies and GC%) to cluster 
    contigs. The published workflow uses this to refine bins 
    initially clustered soley on coverage.
    
    Binsanity-refine can be used as a stand alone script if you don't have more than 2 sample replicates 
    
    """,usage='%(prog)s -f [/path/to/fasta] -l [fastafile] -kmer [kmer type] -c [coverage file] -p [preference] -m [maxiter] -v [convergence iterations] -d [damping factor] -x [contig size] -o [output directory]',formatter_class=RawTextHelpFormatter)
    parser.add_argument("-c", dest="inputCovFile", help="""
    Specify a Coverage File
    """)
    parser.add_argument("-f", dest="inputContigFiles", help="""
    Specify directory containing your contigs
    """)
    parser.add_argument("-p", type=float, dest="preference", default=-25, help="""
    Specify a preference (default is -25)
    Note: decreasing the preference leads to more lumping, 
    increasing will lead to more splitting. If your range
    of coverages are low you will want to decrease the 
    preference, if you have 10 or less replicates increasing
    the preference could benefit you. For complex datasets
    with low abundance organisms a preference
    of -25 was found to be optimal
    """)
    parser.add_argument("-m", type=int, dest="maxiter", default=4000, help="""
    Specify a max number of iterations (default is 4000)
    """)
    parser.add_argument("-v", type=int, dest="conviter",default=400, help="""
    Specify the convergence iteration number (default is 200)
    e.g Number of iterations with no change in the number
    of estimated clusters that stops the convergence.
    """)
    parser.add_argument("-kmer",type=int,dest="inputKmer",default=4,help="""
    Specify a number for kmer calculation. Default is 4. 
    Tetramer frequencies are recommended
    """)
    parser.add_argument("-d",default=0.95, type=float, dest="damp", help="""
    Specify a damping factor between 0.5 and 1, default is 0.9
    """)
    parser.add_argument("-l",dest="fastafile", help="""
    Specify the fasta file containing contigs you want to cluster
    """)
    parser.add_argument("-x",dest="ContigSize", type=int, default=1000,help="""
    Specify the contig size cut-off (Default 1000 bp)
    """)
    parser.add_argument("-o",dest="outputdir", default="BINSANITY-REFINEMENT", help="""
    Give a name to the directory BinSanity results will be output in 
    [Default is 'BINSANITY-REFINEMENT']
    """)
    parser.add_argument('--version', action='version', version='%(prog)s v0.2.5.11')


    args = parser.parse_args()
    if len(sys.argv)<3:
        parser.print_help()    
    if args.inputCovFile is None:
        if (args.inputContigFiles is None) and (args.fastafile is None):
            parser.print_help()
    if (args.inputCovFile is None):
        print "Please indicate -c coverage file"
    if args.inputContigFiles is None:
        print "Please indicate -f directory containing your contigs"
    elif args.inputContigFiles and not args.fastafile:
        parser.error('-l Need to identify file to be clustered')

    else:
        start_time = time.time()
        sys.stdout = Logger()
        print """
        -------------------------------------------------------
                     Calculating GC content
        -------------------------------------------------------"""
        GC_time = time.time()
        GC_Fasta_file(args.inputContigFiles, args.fastafile)
        print "GC content calculated in %s seconds" % (time.time() - GC_time)
        
        print """
        -------------------------------------------------------
                     Calculating tetramer frequencies
        -------------------------------------------------------"""
        kmer_time = time.time()
        output(kmer_counts(args.inputKmer,args.inputContigFiles,args.fastafile))
        print "tertamer frequency calculated in %s seconds" % (time.time()- kmer_time)
        
        print """
        ------------------------------------------------------
          Creating Combined Composition and Coverage Profile
        ------------------------------------------------------"""
        combine_time = time.time()
        print "Combined profile created in %s seconds" % (time.time()- combine_time)
       
        get_contigs(args.inputCovFile)
        print """ -------------------------------------------------------
                Reclustering on Composition and Coverage
        -------------------------------------------------------"""
        print "Preference: " + str(args.preference)
        print "Maximum Iterations: " + str(args.maxiter)
        print "Convergence Iterations: " + str(args.conviter)
        print "Contig Cut-Off: " + str(args.ContigSize)
        print "Damping Factor: " + str(args.damp)
        print "Coverage File: " + str(args.inputCovFile)
        print "Fasta File: " + str(args.fastafile)

        val1, val2 = cov_array((get_cov_data('tetra-GC-out')), args.inputContigFiles, args.fastafile,args.ContigSize)

        print """
        
        -------------------------------------------------------
                        Re-clustering contigs
        -------------------------------------------------------""" 
        affinity_propagation(val1, val2, args.fastafile, args.damp, args.maxiter, args.conviter, args.preference,args.inputContigFiles,args.outputdir)
  
        print("""
        --------------------------------------------------------
              --- Ammended Bins Computed in %s seconds ---
        --------------------------------------------------------""" % (time.time() - start_time))
