Constraint-aware candidate generation that translates biological hypotheses into chemically valid small molecules
or peptide candidates. This service delivers a docking-ready package with explicit traceability
(constraints → candidates → prioritization), designed to connect directly to
Service — Docking and Services — MD Simulation.
Output is designed to be directly consumable by KIOM Stage 2 (Vina/Smina/GNINA) without manual reformatting.
If you already have a curated compound list, consider starting from Service — Docking.
Candidate sets are proposed under explicit constraints defined by a binding pocket, hotspot residues,
pharmacophore patterns, or exclusion rules. The output is curated for docking throughput and downstream ranking.
Sequence candidates for interface blocking or binding enhancement guided by hotspot windows
and screening formats such as peptide arrays.
| Item | Examples |
|---|---|
| Target context | Protein PDB/mmCIF or AlphaFold model; target region (pocket / interface) description |
| Constraints | Hotspot residues, motif to engage/avoid, known actives/inactives (if any), assay constraints |
| Desired output type | Small molecules (SMILES) and/or peptides (FASTA); target count (e.g., 500–10,000) |
| Filters | Drug-likeness, PAINS/toxicophore rules, novelty vs known scaffolds, synthetic feasibility |
If constraints are not yet defined, we can derive a first-pass pocket/hotspot definition from structure.
The objective is not maximum quantity. The objective is a usable candidate set with clear provenance
and direct compatibility with downstream docking and MD validation.
Each step is designed to remain auditable and immediately usable for docking and MD simulation.
Pocket/hotspot definition, target residues, interaction goals, exclusions, and assay constraints.
Propose diverse candidates that satisfy constraints. Optional distribution exploration can be applied
when exploration of novelty is a primary objective.
Apply feasibility and redundancy control, standardize formats, and define the conformer/protonation policy
for docking readiness.
Deliver files and a handoff map that can be executed in Service — Docking and validated in Services — MD Simulation.
Candidate library is formatted to run directly with Vina/Smina/GNINA workflows (Stage 2).
If needed, candidates can be tagged for downstream RF/GNN pipelines after docking (Stage 3).
Top candidates from docking/consensus can proceed to MD protocols as a post-selection validation step.
After Generation, the standard flow is:
Service — Docking (Vina/Smina/GNINA) → Mechanistic validation → Consensus ranking.
MD Simulation is recommended for the final shortlist (Top-N) rather than the full candidate set.
ENGLISH
