What are possible reasons for seeing an IP signal for a protein on a blot, but no corresponding signal for this protein in the input controls run in the same gel?
When you IP the protein, you strongly enrich for it - that is to say, you will have much higher concentration of your protein in the material eluted from your IP, than in your input material. For this reason, most often, the signal on a Western blot is stronger for the IP than for the input. If you cannot see your protein at all in the input material, the reason may be that it is expressed at very low levels, or that your antibody does not have very high affinity for it. You may try to incubate your membrane over night with a higher concentration of primary antibody, which sometimes aids in detection. Also, you may have signal from the light and heavy chain of the antibody you used for the IP (which will not be present in your input), that can interfere/confuse the results. These bands usually run at 25 and 50 kDa, and if you are expecting your protein at around these molecular weights, you need to take care to not misinterpret the blot.
As mentioned by Ronald, I assume your protein is not 25 or 50 kDa for normal antibody-IP (alters for alpaca antibodies). Given that possibility, if your protein is expressed at a very low level then 5% input may not be sufficient to observe a strong signal. Furthermore if the efficiency of your pull down is high then you wouldn't be able to develop the westerns enough to visualize a faint signal (input) in presence of the strong signal (elute). I think simply increasing the input amount given that you can pull down efficiently (and it would also work for slightly smaller volume) would be helpful.
Most likely answer as indicated by others: you are seeing artefact bands from heavy and light chain or from Protein A. Always run a "buffer only" control where you leave out the lysate but go through all of the IP steps. This will tell you which bands were dependent on the lysate. If you can't see a band by doing a straight western blot (which is what the input lane is), try different antibodies and optimize your western blot conditions until this works. Most antibodies do not work, don't be surprised if yours is useless. Then move on to the IP step.