Absorption is as you said, intrinsic and not something can be removed. However, for a certain protein, the absorbance is a certain value at a certain concentration, you might be able to subtract the absorption spectrum of the pure protein as a baseline correction. If it does not work, then maybe think about if any possible fluorescence based assays.

It would help to know what chromophore you are trying to observe. However, if you switch from aqua to DMF you may be able to red-shift the chromophore you want to measure some distance from the aromatics.

Personally I would use electronic spectral subtraction.

Take spectrum of protein: Scan 1

Take spectrum of (dilute) assay mix: Scan 2

Now add assay mix to protein, and then run a few time dependent scans: Scans 3-6.

You can electronically subtract scan 1 from Scan 6 and observe the change, seeing a trough (disappearance of assay mix (Scan 2)) and new peak, your 'new' chromophore.

Details would help

Thanks a lot. I am trying to study the interaction of proteins with various transition metals. However, I am unable to resolve the binding of mixtures of different proteins with metals as all proteins give peak absorbance at 298 nm

What is the wavelength at which you are monitoring the transition metal binding peak? If it is not in the same range as that of protein then you can solve this by taking the same concentration of protein solution (exactly in the same conditions as that of your experimental solution) as a reference instead of pure water.