from collections import defaultdict

kmers = defaultdict(int)

def most_frequent_kmer(seq, k):

    seq = seq.upper()
    assert set(seq).issubset(set(['A', 'C', 'T', 'G']))
    for i in range(0, len(seq)-(k-1)):
        kmers[seq[i:i+k]] += 1
    for key, val in kmers.items():
        if  val == max(kmers.values()):
            print(str(key), end = ' ')

most_frequent_kmer('ACGTTGCATGTCGCATGATGCATGAGAGCT', 4)