Frequently asked questions

We start from the question, not from the data or the method. A short call establishes the decision you need to make, whether it is internal (go or no-go, which candidate, which dose, which design) or regulatory. We then design the MIDD strategy, carry out the analyses ourselves, and deliver results you can act on, documented to submission standard so the same work supports the regulatory file.

If we think modeling is the wrong instrument for your question, we say so at the start, before you have paid for anything.

Within 48 hours of first contact. The proposal sets out the scope, the deliverables, the timeline and a fixed price, so you can decide straight away.

Most of our deliverables are completed within six weeks. The delivery date is fixed in the proposal, and you can follow progress against it at any moment on your project dashboard. We check the state of the data at the start, so the date we commit to is realistic rather than optimistic.

No. We scope the project fully at the start and commit to it: no change orders and no delays. If you later decide to extend the work, that is a new piece of work, agreed separately, never a change to the one already running.

A fixed price, set in the proposal against a defined scope. A focused non-compartmental analysis is a modest, well-bounded piece of work. A full population model with covariate analysis and a submission-ready report is considerably more. Tell us the question and you will have a number, and what it buys, within 48 hours.

Yes. Most of our clients are small and medium-sized companies, and for some it is the first model-informed analysis in the programme. We explain the reasoning in plain terms and hand over something your team can carry forward, rather than a black box that requires us forever.

Yes, as a matter of course. We do not publish client results without explicit permission, which is why the examples on this site are drawn from public guidance and literature rather than from client work.

A documented dataset, the analysis code, the model itself, and a report with the tables and figures, all to submission standard. Everything is generated from code, so it can be re-run, checked and extended. Throughout the project you also have a dedicated web dashboard showing the timeline, the progress and the results as they come in.

Yes. The files are yours. Where we can we favour open tools precisely so that you are not dependent on a licence, or on us, to reopen your own work later.

Yes. That includes the non-compartmental and toxicokinetic analyses, exposure-based safety margins against the toxicology, the human PK prediction, and the starting dose and escalation rationale, written to submission standard. We can also help answer questions that come back.

Raphael Bilgraer leads every engagement and is your point of contact throughout. Beyond that, Quantmetrics staffs each project to what it needs, bringing in pharmacokineticists and clinical pharmacology experts when a programme calls for disciplines beyond one person. You are told who is on the project, and you deal with the people doing the work rather than an account manager.

Either. Some clients want a finished, independent analysis; others want us embedded, joining the calls with their CRO and reviewing protocols before they lock. On longer programmes the combination works best, because the knowledge stays with you.

Small molecules and biologics alike, including monoclonal antibodies, peptides and antibody-drug conjugates. Projects have spanned oncology, immunology, infectious diseases, CNS and metabolic disorders, covering complex pharmacokinetics as well as straightforward profiles: target-mediated drug disposition, non-linear PK, and prolonged-release formulations.

Yes. For complex generics, we use PBPK and physiologically-based biopharmaceutics modeling to run virtual bioequivalence trials, testing the formulation and the study design before the pivotal study is committed. For biosimilars, we design and simulate the PK similarity study, including the sample size, the sampling schedule and the choice of study population.

NONMEM and Monolix for population modeling, PK-Sim/MoBi and Simcyp for physiologically-based models, Phoenix WinNonlin for non-compartmental analysis, and R with its pharmacometrics ecosystem (nlmixr2, rxode2, PKNCA). We also build R-Shiny applications for visualisation and simulation. The choice follows from the question and from what your submission will expect.

Yes. Finished NONMEM runs and Monolix projects can be imported into R and carried forward for simulation, VPC, or conversion to an open workflow. It is also a good way to qualify an existing model independently, which is sometimes what a partner or investor actually wants.

Yes, from first-in-human onward. We deliver the tables, figures and listings (TFLs) for the clinical study report, together with CDISC-compliant ADaM PK datasets, such as ADNCA and ADPP, ready for submission.

Usually not one that prevents the work, though it changes what can be concluded. Sparse sampling still supports a population model, which is part of why the method exists. What matters more is that dosing and sampling times are reliable. If they are not, we will tell you what can and cannot be concluded rather than papering over it.

Both, and usually with the same piece of work. The analysis that answers your internal question, such as which candidate, which dose or which design, is documented to submission standard, so it can go into the regulatory file later without being rebuilt.

Yes. We prepare submission-ready population PK and exposure-response analyses in line with FDA, EMA and PMDA expectations, following current guidance and including the documentation an IND, NDA or MAA filing requires. We can also help prepare briefing documents and responses to questions.

One honest caveat: acceptance depends on the whole dossier and the specific question, not on model quality alone. What we can commit to is that the work is documented, reproducible and presented with its assumptions and limitations stated.

Sometimes, and this is one of the clearest returns on MIDD. Under ICH M12, PBPK can predict drug-drug interactions that were not tested clinically and, in some cases, replace a dedicated interaction study. It only holds when the model is verified against clinical data it was not fitted to, and documented to the standard the agency expects. Where that bar cannot be met, we will tell you the study is still needed.