Entering edit mode
6.5 years ago
Mbillah
▴
140
Recently I read a paper where they estimate their genome size by this formula:
G = k-mer number/k-mer depth
Where, k-mer num=52 413 427 492, K-mer depth = 17
Now, I want to estimate my genome size, where I have 123742407 paired reads and sequence length is 150 bp. if I select kmer as 51. What will be my estimate genome size?
You can't work it out from that alone, til you work out your number of kmers.
Essentially they're taking the number of k-mers of a given length to provide a rough estimate of genome size, since as a genome gets larger, its number of unique k-mers should also increase.
I have found other formula: Estimate genome size
N = M * L / (L - K + 1)N is Depth of Read CoverageM is mean k-mer coverageL is read lengthK is k-mer sizeG = T / NG is the genome sizeT is the total number of basesIn my side here, L=150, k=51 am I right? How can I find N and M?
You have to map your reads, you don’t yet have all the data you need to do this calculation.
http://koke.asrc.kanazawa-u.ac.jp/HOWTO/kmer-genomesize.html
https://bioinformatics.uconn.edu/genome-size-estimation-tutorial/
https://bioinformatics.uconn.edu/genome-size-estimation-tutorial/n = [( L - k ) + 1 ] * Cn = [(150-51)+1]* 123742407= 12374240700N = n / C = 12374240700 / 123742407 = 100Here, sequence length 150, k-mer length 51 and total read 123742407 What is my wrong ?you don't need the number of reads.
what you need is the number of unique kmers in your input data.
You can count those by running software like,
jellyfish,ntCard, ... once you have that count table I would suggest to use the GenomeScope website were all these estimate calculations are done for you.If you want do do the calculation yourself, then you need to plot your kmer count table in a histogram, determine the X-value for the peak in it and use that in your formula
and read the links genomax provided for the detailed guideline of the above
Whee have you got this number from? 123742407