How to control the INCAR tags to get the accurate DOSCAR in interface structure?

[ssaicoda]

Dear all.
I calculate the interface superstructure of metal/Si.
The DOS of bulk structure was calculated.
However, I don't know that DOS of interface in superstructure is accurate or not.
Could you tell me the any tip to control the INCAR tags to get the accurate DOSCAR in interface?

INCAR (using)
IALGO = 48
ISMEAR = -5
SIGMA = 0.050
LREAL = .T.
PREC = High
ISTART = 0
ENCUT = 500
LORBIT = 10
NPAR = 1
NSW = 0
ISIF = 2
IBRION = -1
NFREE = 10
POTIM = 0.5

[admin]

Dear colleague,
to get an accurate DOS with the Tetrahedron method, please use a fine k-mesh which should include the Gamma point and the k-points at the BZ boundary (please check if this is the case).
Also, depending on the energy interval (EMIN,EMAX), it may be necessary to increase the number of grid points at which the DOS is evaluated, especially if the DOS shows sharp peaks.
Please check if NEDOS=301 (=default) has been sufficient in your case. This can easily be done by eg. plotting the DOS together with the integrated DOS: at each energy at which the integrated DOS shows a step, there has to be a peak in the DOS. If this is not the case, please increase NEDOS in INCAR (estimate the required resolution by having a look at the energy dispersion of the KS levels which contribute to sharp peaks, either in OUTCAR or in PROCAR)