should I run SCF calculation separately under various method parameters?

[iamddba]

Hello all!
I want to compare the difference of the zero bias transmission spectrum(TS) under various method parameters, for example K-point samping. The SCF calculation with specific K-point(1x1x500) has been executed and I used it to run the corresponding TS. Then I change K point to another one. Should I re-calculate zero bias SCF in this K-point setup firstly before the new TS calculation will start?
In other words, is the zero bias SCF result in specific parameters always useable for all analysis calculation even one has changed the method parameters?
One more question, in the case above mentioned, should the K-point be samed in "method" with in "analysis" when one calculates TS? I know the post "About the Number of k-points ?" in this forum, but I am still confused a bit.  Which one can bring more effect on the transmission results? 
Best regards!

[zh]

Please note that there are two sets  of k-point  meshes: one k-point mesh for the self-consistent calculation and the other one for the transmission spectrum calculation.
If you just want to check the effect of the former k-point mesh on the transmission spectrum, you may restore the previous converged self-consistent for a smaller k-point mesh and perform the self-consistent calculation for a larger k-point mesh.
If you just want to check the effect of the latter k-point mesh on the transmission spectrum, it is no necessary to perform the self-consistent calculation again and you may restore the previous converged self-consistent calculation to do the TS calculation with a larger k-point mesh.

If you want to check both of them, you may do the first case to get a converge k-point mesh for the self-consistent calculation, and then keep the k-point mesh for the self-consistent calculation and increase the k-point mesh for transmission spectrum calculation.   

Usually, the size of k-point mesh in transmission spectrum calculation is larger than that of self-consistent calculations.

[iamddba]

Thank you, zh. Thanks for your detailed explanation.