Hi,
Given below is an explanation that i found for the different kmer size and its effect on sensitivity and specificity.
"A smaller k-mer value allows locating more overlapping
sequence (i.e. higher sensitivity) while increasing the number of ambiguous repeats(i.e. lower specificity). An extreme example for that is choosing a k-mer length of 3 or 5 where there will be only few possible high coverage k nucleotides with very low specificity. On the other hand, choosing a higher k-mer value has the opposite effect - it decreases sensitivity while increasing specificity.”
From this i understand that an assembly at a higher kmer size is always more "accurate"(not talking about better N50) than the one at a lower kmer size. Is that correct(irrespective of the sequencing errors and polymorphism)?
Could anybody explain how the accuracy of an assembly goes down with lower kmer size?
Thanks
Given below is an explanation that i found for the different kmer size and its effect on sensitivity and specificity.
"A smaller k-mer value allows locating more overlapping
sequence (i.e. higher sensitivity) while increasing the number of ambiguous repeats(i.e. lower specificity). An extreme example for that is choosing a k-mer length of 3 or 5 where there will be only few possible high coverage k nucleotides with very low specificity. On the other hand, choosing a higher k-mer value has the opposite effect - it decreases sensitivity while increasing specificity.”
From this i understand that an assembly at a higher kmer size is always more "accurate"(not talking about better N50) than the one at a lower kmer size. Is that correct(irrespective of the sequencing errors and polymorphism)?
Could anybody explain how the accuracy of an assembly goes down with lower kmer size?
Thanks
Comment