What is the trade-off between promoter strength and specificity?

It is tempting to ask for a promoter that is both maximally strong and perfectly specific, and perhaps compact too. Managing that expectation honestly up front saves disappointment later. Cell-type-restricted elements often (not always) sit below ubiquitous viral promoters on the strength axis, so there is usually a trade-off to reckon with. Strong and specific elements do exist, and a large part of our job is finding or building them for your indication where possible, but also realizing when a balance needs to be met.

The regulatory grammar that drives broad, strong activity is not the same grammar that restricts activity to a single cell type, so specificity often costs some absolute output. That is why the strongest elements, such as viral promoters, tend to be the least selective. But the relationship is a tendency rather than a rule: some genomic elements are naturally both strong and restricted, and an empirical screen across many candidates is how you find those exceptions rather than settling for the average.

In practice we start by defining, together with you, two numbers: the minimum expression your therapy needs to be effective, and the off-target ceiling you can tolerate for safety. A SuRE™ screen then returns the achievable frontier in your own cell type, a set of candidates spanning that strength-and-specificity space, rather than a single take-it-or-leave-it element. That keeps the choice explicit and yours, and it can surface options no one would have picked from sequence alone, sometimes an element that is both strong and selective, and otherwise the best available balance for your requirements.

The trade-off is at the same time not fixed. Synthetic design, increasingly guided by machine learning and AI trained on the data from our screens, can raise strength while holding specificity, effectively pushing the frontier outward. We treat those designs as hypotheses and validate them experimentally rather than trusting the model, so what reaches you is measured, not predicted. Where a genuinely strong and cell-type-specific element is achievable at all, this is often how we get there, and where it is not, it still helps us reach a better balance.

To keep the comparison concrete, we spike well-characterized references into every screen, including cytomegalovirus (CMV), CAG and EF1a, alongside any (tissue-specific) benchmarks you prefer. Candidates are then reported relative to elements you already use, so you can weigh the strength you gain against the specificity you gain, or give up, in terms you already know rather than in the abstract.

Talk to us. Tell us the expression level and specificity your program needs, and we will map the achievable frontier in your cell type, including whether a strong and selective element is within reach.