1. If your proteins of interest have already bound to the matrix at the given conditions (depending when you mixed and what you did), just use a normal column to hold the matrix/ beads and wash several times with PBS etc. to get rid of non-specific proteins/ metabolites/ carbohydrates etc. Following this, use elution buffer (depending on what kind of interaction you have facilitated for the protein with matrix) to elute your proteins selectively.

2. Another alternatively, wash column with cellulases, DNAases, RNAases etc. (except peptidase/ proteases) and in presence of protein protecting agents like PMSF etc. to wash off everything unwanted, followed by solvent washes, so that only beads + proteins of interest remains: and this can be further centrifugation-washed to collect beads + proteins and may be separated in acetone/ ethanol washes.

3. Use Centricons/ Amicons for a protein range of interest and a better cut-off so that everything else either stays back or flows through. Beads may not be recovered but proteins for sure will be !

These above all are theoretical suggestions- and a good book like Sambrook Russell etc. must be consulted before proceeding with exact steps, buffers, chemical compositions etc. These are only clues, as far as I could understand your  issue.

Thanks,

Biswa

I used the Ni resin for His-tag purification