You’re welcome Aasif. That is exactly the point, and this should always be considered in all system designs. Otherwise there should have been a paper introducing the best scheme in a general sense, and not only for a specific system (which in the case of the aforementioned paper is NRZ for 10 Gbps).
Regards