The modern PB/PK approach
Revolutionise your PB/PK!
By combining cutting edge ML and AI techniques
with traditional compartment models
we can help solve your PB/PK nightmares.
Scroll
Every square represents a drug going to trial in a given year1
These drugs failed trials
These drugs failed trials directly because of bad PB/PK2
These drugs failed for reasons downstream of PB/PK3
Each one of these squares
represents $100M USD
of spending.4
It doesn't have to be this way.
- Successful candidates
- Failed: PB/PK
- Downstream of PB/PK
- Other failures
Scroll
Scroll
At Informed Kinetics, we use cutting edge ML and AI to drastically improve your PBPK
Footnotes and citations
- 1. Derived from CDER 2025 novel approvals (46) and the industry-wide ~6.7% Phase-I-to-approval likelihood reported in Citeline / Norstella. Steady-state approximation — drugs approved in a given year actually entered Phase I 8–10 years earlier. Norstella / Citeline
- 2. Sun D. et al., 2022, “Why 90% of clinical drug development fails and how to improve it.” Acta Pharmaceutica Sinica B. Sun et al., 2022
- 3. Internal estimate combining exposure-driven toxicity and target-exposure efficacy failures; treated as the high-end of the PK-attributable range. Not a published figure.
- 4. ASPE, “Examination of Clinical Trial Costs and Barriers to Drug Development”; corroborated by capitalised R&D-per-approval figures in Wouters et al., 2020, JAMA (~$985M ÷ ~10 entrants). ASPE
- 5. The ADME screening that reduced PK failure from ~40% to ~11% in small molecules relies on heuristics (e.g. Lipinski’s rule of five) that do not apply to nanoparticles, so the failure rate regresses toward the pre-screening era. Argued from Sun et al., 2022; no published NP-specific figure exists. Sun et al., 2022
- 6. “Pharmacokinetics and tumor delivery of nanoparticles,” PMC10686544. NP behaviour is dominated by biodistribution and immune clearance; almost all PK/PD-governed. PMC10686544