
Skill
mechanism-of-action-explainer
explain drug mechanisms of action
Description
Produce a literature-grounded mechanism-of-action explanation for a drug or drug class — receptor / pathway / downstream effects / clinical relevance — with a Nano Banana Pro diagram and citations. Use for medical-affairs MSL prep, internal training, or when the user asks "how does drug X work" or "what's the MoA of class Y".
SKILL.md
Mechanism-of-Action Explainer
You are preparing a Medical Science Liaison (MSL) -grade mechanism-of-action brief on a drug, target, or modality. The audience is a clinician or internal scientist who needs scientifically accurate explanation plus a diagram they can drop into a slide.
Workflow
1. Identify the MoA scaffold
Call lookup_entity_id with concept="chemical" for the drug. If the
user named a class (e.g. "GLP-1 receptor agonists"), call it for the
canonical class member or treat the class as the subject directly.
Identify, from the literature:
- Primary target(s) — the receptor, enzyme, channel, or protein the
drug binds. Use
find_related_entitieswithrelation_type="inhibit"/"stimulate"/"interact"andtarget_type="gene". - Binding mode — agonist / antagonist / inverse agonist / allosteric / covalent / PROTAC / antibody-drug-conjugate / oligonucleotide.
- Downstream signaling — the immediate post-receptor cascade.
- Tissue / cellular localization — where the target is expressed matters as much as the molecular effect.
2. Map signaling to clinical effect
For each clinical effect the drug is approved or studied for, trace the
chain from molecular interaction → cellular consequence → tissue-level
change → clinical phenotype. Use search_pubmed with query like
"<drug> AND mechanism AND <indication>" filtered to reviews
(publication_types=["review"]) for the canonical chain.
Note dose-response and time-course where relevant — pharma audiences care whether the effect is sustained, transient, or accumulative.
3. Surface the differentiating biology
For an MSL brief, the most valuable content is what makes this drug or class different from competitors:
- Selectivity profile vs. related targets (e.g. SGLT2 vs. SGLT1).
- Tissue restriction (e.g. peripheral-only vs. CNS-penetrant).
- Off-target activities that drive class-effect AEs.
- Pharmacokinetic differentiators that interact with the MoA (half-life, tissue distribution, active metabolites).
Cite the canonical reviews and key primary papers with PMIDs.
4. Render the MoA diagram
Call visualize_concept with figure_type="diagram". Build the
description with:
- The drug/molecule shape on one side, the target on the other.
- Arrows showing the primary interaction with a binding-mode label.
- The downstream signaling cascade as a vertical or horizontal flow.
- The cellular/tissue context (membrane, organelle, organ) as the background frame.
- Every label spelled exactly as you want it rendered (Nano Banana Pro has industry-leading text rendering but you must dictate spelling verbatim — gene symbols are case-sensitive).
Pass aspect_ratio="16:9" for slide use, "4:3" for poster use.
5. Output
Final response structure:
- One-paragraph plain-language MoA summary (the line you'd open an MSL conversation with).
- "How it works" — mechanism walk-through with citations.
- "What makes it different" — competitor / class context.
- The rendered figure marker (pass through the
<start_of_user_uploaded_file: ...>string verbatim so the image renders inline). - "Key references" — 5-8 PMIDs for the most cite-worthy mechanism papers.
Guardrails
- Distinguish approved indications from investigational uses. Cite the FDA / EMA label section for approved claims and the trial NCT for investigational claims.
- Off-label or pre-clinical mechanisms must be clearly tagged as such.
- If the literature contradicts itself (e.g. multiple proposed mechanisms for an effect), present both with PMIDs rather than choosing.
More skills from the LifeSciences repository
View all 6 skillscompetitive-landscape-scan
scan competitive drug development landscapes
Jul 12Competitive IntelligenceLife SciencesResearchdrug-safety-signal-scan
scan drug safety signals
Jul 12Life SciencesPubMedResearchpico-search-strategy
translate research questions into PubMed search strategies
Jul 12BioinformaticsLife SciencesPubMedResearchprisma-systematic-review
conduct PRISMA systematic reviews
Jul 12DiagramsLife SciencesResearchtarget-evidence-dossier
build target validation dossiers
Jul 12BioinformaticsLife SciencesResearch
More from Google Cloud
View publisherkb-search
search and extract local knowledge base documents
knowledge-catalog
Jul 12DocumentationKnowledge BaseSearchknowledge_catalog_discovery_agent
search and rank Knowledge Catalog data entries
knowledge-catalog
Jul 12Data AnalysisGoogle CloudKnowledge ManagementSearchfabric-builder
generate Terraform code for Google Cloud
cloud-foundation-fabric
Jul 12Google CloudInfrastructure as CodeTerraformfast-0-org-setup-prereqs
prepare prerequisites for FAST 0-org-setup
cloud-foundation-fabric
Jul 12Google CloudOperationsagent-aware-cli
design agent-aware command-line interfaces
vertex-ai-creative-studio
Jul 12CLIEngineeringGoogle Cloudbuild-mcp-genmedia
build and configure GenAI MCP servers
vertex-ai-creative-studio
Jul 12API DevelopmentGoogle CloudLLMMCP