FROGENT / capability index
Eleven scientific tasks. Eleven evidence-bearing pages.
Open the task you need. Each page explains the scientific question, the agent action, its inputs and outputs, the source figure, and the boundary of the evidence.
What evidence should shape this scientific question, and which sources support it?
↗02File acquisitionCan the workflow acquire the requested scientific file and keep its identity and provenance attached?
↗03Target identificationWhich disease-associated targets should be prioritized for a computational discovery program?
↗04ADMET predictionWhich predicted developability properties or liabilities should inform the next candidate decision?
↗05Small-molecule dockingHow do proposed small molecules pose within a prepared pocket under one stated docking protocol?
↗06Peptide dockingHow does a peptide candidate pose against a target under a declared in silico protocol?
↗07Interaction analysisWhich pose-level contacts support or challenge a proposed protein–ligand binding hypothesis?
↗08Lead optimizationHow should a seed candidate change when predicted properties and structural diagnostics are reviewed together?
↗09Molecule generationWhich new molecular candidates satisfy the stated pocket, fragment, and design constraints?
↗10Synthesis planningWhat retrosynthetic route could connect a computational candidate to candidate starting materials?
↗11Report writingHow can the complete computational campaign be assembled into one reviewable scientific record?
↗Every capability reads from and writes to the same campaign record.
Experts can inspect evidence, constraints, rankings, failures, and artifacts before the next decision.
Predictions and generated proposals stay distinct from experimental or clinical confirmation.