背景
Kmer是基因组组装算法中经常接触到的概念,简单来说,Kmer就是长度为k的核苷酸序列。一般长短为m的reads可以分成m-k+1个Kmer。Kmer的长度和阈值直接影响到组装的效果。
Denovo组装流程:原始数据——数据过滤——纠错——kmer分析——denovo组装
。
组装测序策略:根据基因组大小和具体情况选择个大概的k值,构建contig所需的数据量以及所需的构建的文库数量。对于植物基因组一般考虑的是大kmer(>31),动物一般在27左右,具体根据基因组情况调整。需要在短片段数据量达到20X左右的时候进行kmer分析。Kmer分析正常后,继续加测数据以达到最后期望的数据量。
编码
import os
import sys
# convert command line arguments to variables
kmer_size = int(sys.argv[1])
count_cutoff = int(sys.argv[2])
# define the function to split dna
def split_dna(dna, kmer_size):
kmers = []
for start in range(0,len(dna)-(kmer_size-1),1):
kmer = dna[start:start+kmer_size]
kmers.append(kmer)
return kmers
# create an empty dictionary to hold the counts
kmer_counts = {}
# process each file with the right name
for file_name in os.listdir("."):
if file_name.endswith(".dna"):
dna_file = open(file_name)
# process each DNA sequence in a file
for line in dna_file:
dna = line.rstrip("\n")
# increase the count for each k-mer that we find
for kmer in split_dna(dna, kmer_size):
current_count = kmer_counts.get(kmer, 0)
new_count = current_count + 1
kmer_counts[kmer] = new_count
# print k-mers whose counts are above the cutoff
for kmer, count in kmer_counts.items():
if count > count_cutoff:
print(kmer + " : " + str(count))
网友评论