Simple macrocyclic water-soluble hosts such as cucurbiturils, cyclophanes, and calixarenes have long been used for biosensing via indicator displacement assays. Using multiple hosts and dyes in an arrayed format allows pattern recognition-based "chemical nose" sensing, which confers exquisite selectivity, even rivaling the abilities of biological recognition tools such as antibodies. However, a challenge in indicator displacement-based biosensing with macrocyclic hosts is that selectivity and scope are often inversely correlated: strong selectivity for a specific target can limit wide application, and broad scope sensing can suffer from a lack of selectivity between similar targets. This problem can be addressed by using water-soluble, self-folding deep cavitands as hosts. These flexible bowl-shaped receptors can be easily functionalized with different motifs at the upper and lower rim, and the large cavities can bind many different fluorescent dyes, causing either fluorescence enhancement or quenching upon binding.Cavity-based affinity is strongest for NMe3+ groups such as trimethyl-lysine, and we have exploited this for the site-selective recognition of post-translational lysine methylations in oligopeptides. The host recognizes the NMe3+ group, and by applying differently functionalized hosts in an arrayed format, discrimination between identical modifications at different positions on the oligopeptide is possible. Multiple recognition elements can be exploited for selectivity, including a defined, yet "breathable" cavity, and variable upper rim functions oriented toward the target.While the performance of the host/guest sensing system is impressive for lysine methylations, the most important advance is the use of multiple different sensing mechanisms that can target a broad range of different biorelevant species. The amphiphilic deep cavitands can both bind fluorescent dyes and interact with charged biomolecules. These non-cavity-based interactions, when paired with additives such as heavy metal ions, modulate fluorescence response in an indirect manner, and these different mechanisms allow selective recognition of serine phosphorylation, lysine acetylation, and arginine citrullination. Other targets include heavy metals, drugs of abuse, and protein isoforms. Furthermore, the hosts can be applied in supramolecular tandem assays of enzyme function: the broad scope allows analysis of such different enzymes as chromatin writers/erasers, kinases, and phosphatases, all from a single host scaffold. Finally, the indirect sensing concept allows application in sensing different oligonucleotide secondary structures, including G-quadruplexes, hairpins, triplexes, and i-motifs. Discrimination between DNA strands with highly similar structures such as G-quadruplex strands with bulges and vacancies can be achieved. Instead of relying on a single highly specific fluorescent probe, the synthetic hosts tune the fluorophore-DNA interaction, introducing multiple recognition equilibria that modulate the fluorescence signal. By applying machine learning algorithms, a classification model can be established that can accurately predict the folding state of unknown sequences. Overall, the unique recognition profile of self-folded deep cavitands provides a powerful, yet simple sensing platform, one that can be easily tuned for a wide scope of biorelevant targets, in complex biological media, without sacrificing selectivity in the recognition.