Why longer K-mer is better for de novo assembly of high abundant transcript and vice versa?

Question

[seta]

Hi all friends,

I know that the high abundant transcripts are better assembled and represented by longer k-mer and shorter k-mer is suitable for reconstructing the low abundant transcript, but I'm not sure about the exact reason of it. Could you please explain to me why?

Thanks a lot

Answer 1

[iraun]

Having a high quantity of expression, i.e high abundance of transcript, it could be possible to assembly them using a long k-mer instead of a shorter one. I mean, the basic idea about the k-mers is:

The shorter k-mer, the bigger sensitivity but smaller specificity: you will be able to assembly more reads (kmers) together but with larger error rate (i.e. you will assembly k-mers together that are not true).

Sohaving a high amount of data, you have more chance to assembly the correct kmers together and you could specify larger k-mers in order to reduce the error rate.

I hope you find this information useful.