Shifts of initial point in plastid assembly in comparison to the reference depending on k-mer length

[AndriiTarieiev]

Dear developers,

I'm experiencing significant shifts of initial/starting point in comparison to the reference sequence when doing plastid assembly using NOVOPlasty. These shifts are dependent on k-mer length. I would appreciate it very much if you are able to suggest any solution to this problem.

Best regards,
Andrii

[ndierckx]

You mean you are getting different assemblies depending on the given k-mer?

[AndriiTarieiev]

No, the assemblies with different k-mers are actually very similar. However, when I align them between each other and to the reference, there is a shift of initial/starting position (where the circular genome is split to linearize it) on ~23-75kbp (shift increases with increasing k-mer size).

[AndriiTarieiev]

To make the problem a bit more visual, I prepared the following illustration based on one real sample:
NOVOPlasty_plastid_assembly_&_alignment.pdf

[ndierckx]

Hi, I do not understand the issue, does it matter where it starts, it is a circular genome, there is no starting point. Starting point will depend on at which time in the assembly process both ends connect with each other.