Yes, there are quite a few groups who have focused on fully automated sample preparation and isolation using acoustofluidics, cell rupture and lysis via focused ultrasonics, and biomolecular separation using a variety of instrumental methods including LC, nanoLC, CE, GC, MST, and various permutations and derivatives have emerged in the search for the optimal combination of methods. That is not to mention the final ToF/ToF, MS/MS, Orbitrap, or another destructive analysis, partitioning samples into destructive and non-destructive analyses, as well as autonomous and automated data processing, analytics, and visualization.

The general consensus is that the choke point in these analyses exist at the point of sample separation, microfluidic devices have proven to be capable of replacing single sample runs with synchronized iterative sample separation that can take advantage of the mass spectrometers extremely fast and sensitive detection that is currently under ~0.01% utilization. The way through is, in this researchers opinion, when the computer scientists and software engineers get together with the bioanalytical chemists in order to focus on high-throughput, fully-automated, separation of targets of interest from blood (from exosomes to circulating tumor cells), automated online sample preparation (ultrasonic or optical tweezing), and automatic online separation, partitioning, identification, quantitative, and characterization (CE and MST will undoubtedly replace LC). We're still missing control systems and algorithms to coordinate this orchestra, but the fundamental question of plausibility has been demonstrated by many groups to be a resounding yes. The technology exists, it works, we just haven't been able to tie it all together yet.