What is the preclinical stage in drug development?
Preclinical work (also called nonclinical) is the bridge between a lead molecule that looked good in a screen and a drug you can defend to the FDA in an IND. It answers three blunt questions about your candidate before any human takes it: does it actually do what you think it does in a living system, what does the body do to it once it is dosed, and what is it likely to break.
Discovery hands you a molecule with attractive in vitro potency. Preclinical pressure-tests that promise. You confirm the mechanism carries from a biochemical assay to cells to a whole animal, you measure exposure (how much drug reaches the blood and the target tissue over time), and you take a first look at safety in non-GLP exploratory studies. The output is a coherent story: target engagement, efficacy in a relevant model, a clean enough PK profile to pick a dose, and no showstopping liability that would sink the program.
One distinction is worth keeping straight. The exploratory studies in this stage are usually non-GLP and meant to inform decisions, not to file. The pivotal, GLP-compliant safety package (repeat-dose toxicology in two species, the safety pharmacology core battery, genetic toxicology) belongs to the IND-enabling stage that follows. Treating early preclinical as decision-grade science rather than regulatory paperwork is what keeps the budget sane. You kill weak candidates here, cheaply, before you commit to an expensive GLP tox program.
What CRO and CDMO services do buyers source at the preclinical stage?
Almost everything at this stage is CRO work (lab studies, not manufacturing), though bioanalytical straddles the line because the same assay groups support nonclinical and clinical sample analysis. The categories below map to how sponsors actually carve up the work and write a statement of work.
- In vitro pharmacology: target engagement and selectivity in biochemical and cell-based assays, dose-response (IC50/EC50), receptor binding and functional panels, off-target screening, and cardiac liability flags such as a hERG patch-clamp or automated ion-channel panel.
- In vivo pharmacology and efficacy: proof of concept in disease-relevant animal models (xenograft, syngeneic, PDX, genetically engineered, or humanized for immuno-oncology), tumor growth inhibition or functional readouts, and dose-response in the model that matters for your indication.
- DMPK / ADME: metabolic stability in liver microsomes and hepatocytes, CYP inhibition and induction, plasma protein binding, Caco-2 or MDCK permeability, transporter assays, in vivo PK across species, bioavailability, and metabolite identification to catch reactive or major circulating metabolites early.
- PK/PD and modeling: connecting exposure to effect, building PK/PD models, allometric scaling to project a human dose, and supporting first-in-human dose selection, often with PBPK modeling.
- Bioanalytical: LC-MS/MS or ligand-binding (ELISA, MSD) method development and validation to quantify drug and metabolites in plasma and tissue, with fit-for-purpose or GLP-validated methods depending on whether the data is exploratory or pivotal.
- Biomarker discovery: identifying and qualifying pharmacodynamic, predictive, or target-engagement biomarkers, plus assay development (flow cytometry, IHC, qPCR, multiplex) to carry into the clinic.
- In vitro and early toxicology: exploratory cytotoxicity, cardiotoxicity and hepatotoxicity screens, early genotoxicity flags (Ames, micronucleus), and dose-range-finding (DRF) studies that set the doses for the later GLP tox program.
How do you choose a preclinical CRO?
The first filter is fit, not price. A site that runs flawless small-molecule DMPK is the wrong choice for an antibody or an AAV gene therapy, where the live question is biodistribution and immunogenicity, not CYP inhibition. Match the supplier to your modality (small molecule, biologic, ADC, oligonucleotide, cell or gene therapy) and your therapeutic area before you look at a quote, because the assay menu and the right animal models change completely across those.
Next is the model and method match. For in vivo efficacy, ask whether they already have the specific disease model running and validated in-house, with historical control data, rather than spinning one up for the first time on your dollar. For DMPK and bioanalytical, ask which species, which matrices, and whether they can hit the lower limit of quantification a low-dose program needs. Vague capability claims are cheap. A method that already detects your compound at the concentration you care about is not.
Then look at the practical things that decide whether a program runs on time: capacity and current queue (a great lab booked solid for four months may be slower than a good lab with an open slot), turnaround on the report and not just the bench work, how they price and handle a change order when a study slips or a cohort fails, and the quality of the scientist you will actually talk to. Data integrity and transparent reporting matter as much as headline price. A cheap study you cannot use, or cannot reconcile, is the most expensive outcome there is.
How long does preclinical testing take and what does it cost?
Timelines turn on how much work you are stacking and how serial it has to be. A single in vitro pharmacology or DMPK assay panel can turn around in a few weeks. An in vivo efficacy study runs longer once you account for animal acclimation, dosing, the readout window, and tissue analysis, so plan in months rather than weeks for the full readout. Bioanalytical method development and validation is often the hidden critical path, because in vivo PK and tox samples cannot be analyzed until the method exists.
The honest answer on cost is that it depends on scope, species, and modality, and any supplier who quotes a flat number before seeing your statement of work is guessing. What is reliable is the shape of the spend: in vitro and DMPK panels are the cheapest per study, in vivo efficacy and dose-range-finding sit in the middle, and the work scales up sharply once you cross into the GLP IND-enabling program. The discipline that saves real money is sequencing. Run the cheap, decision-grade studies first, retire weak candidates early, and only fund the expensive pivotal work once the candidate has earned it.
One sequencing note that trips up first-time sponsors: bioanalytical and DMPK are dependencies for almost everything downstream, so getting those suppliers locked and methods built early keeps your in vivo and tox timelines from sliding while samples sit in a freezer waiting on an assay.
Do preclinical studies need to be GLP compliant?
Not all of them, and this is where money gets wasted in both directions. Good Laboratory Practice (21 CFR Part 58 in the US, OECD GLP for ex-US work) governs the safety studies that support an IND. The pivotal toxicology that goes into your regulatory package must be GLP. But the exploratory pharmacology, the early DMPK screens, the dose-range-finding studies, and the proof-of-concept efficacy work in this stage are typically non-GLP, because their job is to inform your decisions, not to satisfy a regulator.
The trap is reading that as a license to be sloppy. Non-GLP does not mean low quality. It means the study is for internal decision-making rather than regulatory submission. The data still has to be solid enough to bet a program on, and a bioanalytical method built fit-for-purpose now is often the same method you validate to GLP later. The opposite trap is gold-plating: paying GLP premiums and accepting GLP timelines for an exploratory screen nobody will ever file. Know which bucket each study sits in before you sign the statement of work.
GCP and GMP do not really apply yet at this stage, since those govern clinical conduct and manufacturing. What you do want to confirm is the CRO's broader quality posture: documented SOPs, data integrity practices, appropriate animal welfare accreditation such as AAALAC for in vivo work, and a clear paper trail, so the non-GLP data holds up internally and the eventual GLP work runs in a system you already trust.
How does sourcing preclinical services through BioBridgeX work?
BioBridgeX is a neutral marketplace, not a lab. It owns no benches and runs no studies, so there is no quiet incentive to steer your DMPK package or your in vivo efficacy work toward a preferred site to fill in-house capacity. Its only job is to match you to qualified CROs and keep the project moving.
You describe the work (the assays, the species, the modality, the indication, the timeline) and get matched with qualified preclinical CRO suppliers. You compare them on capability, transparent quotes, and turnaround in one view instead of running a separate sales cycle with each one. Supplier profiles are public, so you can see who covers in vitro pharmacology, DMPK/ADME, bioanalytical, or biomarker work before you ever talk to a salesperson.
The commercial model is deliberately plain. BioBridgeX is free for buyers. Suppliers pay a flat 2% platform fee, the same rate whether the engagement is a small bioanalytical method transfer or a multi-study pharmacology and DMPK package. When you split work across several suppliers, you compare structured quotes and contract directly with each supplier you choose, keeping sourcing and comparison in one place. The coverage spans every indication and modality, so the same path carries you from preclinical into IND-enabling and beyond without re-papering a new master agreement each time you change suppliers.