Disease-to-target search
Gather supported target hypotheses from disease or phenotype context while keeping source evidence and unresolved queries explicit.
Target intelligence and selectivity
Resolve the biological question before ranking molecules.
Search from a disease, phenotype, natural source, gene, or protein; separate exact target identity from fallback search; and connect tractability and structure readiness with molecule-level off-target evidence.
Target discoveryDisease associationUniProt identityTractabilityOff-target predictionSelectivity
Evidence before target selection
Make identity, relevance, tractability, and selectivity separate decisions.
MolexIO distinguishes a search result from an exact biological entity and a target hypothesis from a validated mechanism. Source records, identifiers, evidence categories, structures, pockets, and model readiness remain visible as a program moves from target discovery into molecule design.
Identity resolution
Reconcile gene symbols, names, accessions, species, and relevant subunits before assigning structures or activity data.
Evidence categories
Inspect genetic, pathway, literature, pharmacology, and other available evidence as distinct signals rather than one unexplained number.
Tractability and readiness
Review known ligands, structure availability, pockets, and supported computational capabilities for the exact target.
Off-target profiles
Screen a molecule against a documented target panel and retain per-target method, score, evidence, and coverage.
Selectivity context
Compare intended-target and off-target signals without labeling missing or incomparable results as demonstrated selectivity.
From context to program target
Preserve the reason a target entered the project.
- 01
Frame the question
Start from disease, phenotype, gene, protein, natural source, or known molecule context.
- 02
Resolve identity
Confirm the exact biological entity and flag ambiguous or unresolved results before reuse.
- 03
Assess readiness
Review evidence, known chemistry, structures, pockets, and model capability for the target.
- 04
Challenge selectivity
Evaluate molecule-level off-target risk and define the experimental panel needed next.
Scientific boundary
Target association is not target validation, and prediction is not selectivity.
Disease associations can be indirect, context dependent, or affected by source bias. Off-target models depend on training coverage, molecular applicability, and panel composition. MolexIO reports unresolved identity, missing coverage, and prediction status so teams do not mistake a search hit or computational margin for causal biology or measured selectivity.
Questions before use
Target and selectivity FAQ
Does target discovery identify a validated target?
No. It ranks hypotheses from available evidence; biological and therapeutic validation remain separate work.
Is off-target prediction an experimental selectivity panel?
No. It prioritizes possible risk and follow-up but does not replace measured binding or functional data.
Why resolve the exact target identity?
Isoforms, species, complexes, and accessions can change the relevant biology, structures, and evidence.
Connect the target hypothesis to every downstream decision.
Keep biological context, structural readiness, compound evidence, and selectivity review attached to the same program.