It depends on the question you need to answer. For example for the patients that I follow, protein based biomarkers (or sometimes functional assays looking for the functionality of these proteins) are fundamental for the phenotypic diagnosis of a disease, and genotypic markers are important for genetic counseling.
Moreover, some genes present mutations affecting different domains of the related protein causing astonishing different phenotypes (e.g.: STAT1 loss of function mutations, that can be recessive or dominant with susceptibility to different pathogens, and gain of function mutations with other different phenotypes).
So, for some diseases the genotypic-phenotypic correlations are not so direct and both assays are important to understand the biological phenomena.
generally proteins can be easily assessed by immunohistochemistry. This makes them easily applicable in routine diagnostics. DNA is more laborious and costy, but the results are harder in terms of accuracy. It depends on the thing you want to monitor with your biomarker.
Profiles of DNA markers (e.g. somatic mutations in DNA escaping tumours) or protein biomarkers (altered signalling, cytokines, growth factors etc) in peripheral blood can be useful for some conditions. For ready-made panels of protein biomarkers, check out for instance the multiplex protein 96 sample assays from OLINK Bioscience http://www.olink.com/products/proseek-multiplex.
Protein and gene combination would be ideal but often gene and protein profiles may not be the same due to various protein half-life. I generally found gene biomarkers more reliable however more expensive to assay than protein biomarkers.
It depends on the question you need to answer. For example for the patients that I follow, protein based biomarkers (or sometimes functional assays looking for the functionality of these proteins) are fundamental for the phenotypic diagnosis of a disease, and genotypic markers are important for genetic counseling.
Moreover, some genes present mutations affecting different domains of the related protein causing astonishing different phenotypes (e.g.: STAT1 loss of function mutations, that can be recessive or dominant with susceptibility to different pathogens, and gain of function mutations with other different phenotypes).
So, for some diseases the genotypic-phenotypic correlations are not so direct and both assays are important to understand the biological phenomena.
Politically to say both of them are important and should be studied to make the full story. I am a believer that regardless how DNA, RNA or even epigenomic change without change in protein or other end product then the story is incomplete. At least my DNA will be the same when I am healthy or ill, maybe not after radiation injury, but my protein profile will change when I am suffering from disease.
Using protein based markers can give you a fine, specific and precise representation of the conditions under consideration. For eg., in case of any infection or any physiological disease, only effector molecules, i.e. proteins, will change, and NOT the DNA of the host, as a result of those conditions. But, DNA markers for any conditions, if available& reliable, are more robust and reproducible unlike protein bio-markers which are quite sensitive to the slight perturbations.