Explicit structure changes
Use aligned 2D and 3D views to distinguish the retained scaffold from atoms removed, changed, or introduced in the candidate.
ADMET-aware lead optimization
See what changed before choosing the next analog.
Compare a seed molecule with generated or selected candidates, inspect the exact chemical transformation, and evaluate ADMET, safety, potency, and makeability trade-offs without collapsing every decision into one opaque score.
Matched molecular changesADMET comparisonPareto analysisMPOStructural alertsEvidence coverage
From liability to testable chemistry
Connect a proposed structural change to the properties it may influence.
MolexIO keeps the seed, candidate, transformation, endpoint values, uncertainty, and evidence status together. A highlighted structural alert is treated as a rule match—not as a measured toxic region—and inherited alerts remain visible when a candidate is compared with its parent.
Seed-versus-candidate ADMET
Overlay both molecules on one radar and inspect endpoint values and deltas beside the chart rather than inferring change from color alone.
MMP and analog hypotheses
Review matched molecular pair transformations and generated analog proposals as chemistry hypotheses with their source and operator retained.
Pareto trade-offs
Identify candidates that preserve meaningful alternatives across selected objectives instead of forcing a premature universal ranking.
Safety gates with context
Separate inherited structural warnings from newly introduced liabilities and distinguish advisory filters from experimental toxicity evidence.
Reusable candidates
Promote a reviewed candidate into the project library so the same molecular identity can enter downstream screening and design workflows.
Target optionalPure ADMET optimization does not require a protein target.
Comparable endpointsSeed and candidate predictions retain the same endpoint definitions and engine context.
No silent evidenceHotspots, poses, or atom contributions are shown as unavailable when they were not calculated.
A reviewable optimization loop
Move from a seed to a short list without losing the reason for each change.
- 01
Choose the seed
Select a project molecule or submit a structure and establish the endpoints that matter for the current decision.
- 02
Generate or compare
Produce candidate hypotheses or bring existing analogs, preserving transformation and parentage.
- 03
Inspect trade-offs
Review structural differences, ADMET deltas, safety warnings, uncertainty, and Pareto membership.
- 04
Promote and test
Add selected chemistry to the library and define the experimental or computational evidence needed next.
Scientific boundary
Optimization proposes priorities, not proven improvements.
A favorable predicted delta can be sensitive to model applicability, molecular representation, endpoint definition, and uncertainty. Pareto status means a candidate represents a non-dominated trade-off under the selected objectives; it does not prove efficacy, safety, exposure, or synthetic success. MolexIO keeps these boundaries visible so teams can decide what to synthesize and measure next.
Questions before use
ADMET optimization FAQ
Does ADMET optimization prove that an analog is safer?
No. It helps prioritize compounds and liabilities for follow-up; experimental safety evidence is still required.
Is a biological target required?
No. Add target context only when the objective also includes potency, docking, or selectivity.
What does Pareto optimal mean?
The candidate is not clearly worse than another option across every selected objective. It is a trade-off candidate, not a guaranteed winner.
Review the chemistry and the property trade-off together.
Use MolexIO to move a defensible candidate from an ADMET comparison into the next project workflow.