Modify the Jukes-Cantor function to compute the distance matrix for 
genome sequences. For a pair of genomes x and y we define the 
evolutionary distance to be
       
            (           Unique common kmers between x and y        )
d(x,y) = -ln(                   --------------------               )
            (  Min(total unique kmers in x, total unique kmers in y)


You will also need a modified version of the function fast_blast_score written in the 
alignment library. Both original functions are given below.

Submit hardcopies in class on Apr 15th.

def distance_matrix(data, D):
    for i in range(1, len(data), 2):
        l = []
        for j in range(1, len(data), 2):
            seq1 = data[i]
            seq2 = data[j]
            pab = 0
            for k in range(0, len(seq1), 1):
                if seq1[k] != seq2[k]:
                    pab += 1
            pab = pab/len(seq1)
            l.append(-3/4*math.log(1-4/3*pab))
        D.append(l)
    return            

def fast_blast_score(seq1,seq2):

    kmerlength = 30
    dseq1 = {}
    dseq2 = {}
    
    for i in range(0, len(seq1)-kmerlength+1, 1):
        kmer = seq1[i:i+kmerlength]
        dseq1[kmer] = 1

    for i in range(0, len(seq2)-kmerlength+1, 1):
        kmer = seq2[i:i+kmerlength]
        dseq2[kmer] = 1

    k = list(dseq1.keys())

    count = 0
    for i in range(0, len(k), 1):
        if k[i] in dseq2:
            count += 1
    

    return count