Curator state

Recently surfaced

  • Screen a polypharmacy medication list for drug-drug interactions (added 2026-06-14) — rung-2 DDInter skill recipe: per-drug DDInter ID resolution → pairwise interaction queries → cited severity/mechanism/management table with explicit “clean” lines, plus an optional rung-3 DailyMed + ClinPGx overlay on major pairs (phenoconversion bridge to the PGx dosing recipe). Drug Repurposing and Discovery focus-day recipe; cookbook’s first DDI-screening recipe. Reported — Domián et al., Explor. Res. Clin. Soc. Pharm. 2025 shows ungrounded LLMs over-flag/hallucinate DDIs (Copilot 1,813 vs a 204-interaction reference on 57 patients), so screening must be database-anchored; DDInter is the peer-reviewed reference (Xiong et al., NAR 2022).
  • Profile a cancer cohort’s genomics with cBioPortal (added 2026-06-14) — rung-2 cBioPortal skill recipe: study/molecular-profile lookup → per-gene mutation+CNA alteration frequency + co-occurrence/mutual-exclusivity → TMB summary → Kaplan-Meier OS split by mutation status, with cohort-denominator caveats enforced. Translational Medicine focus-day recipe; cookbook’s first cohort-level cancer-genomics recipe, cross-linked to the gene-centric target dossier, single-variant variant-interpretation, and adjusted-modelling survival recipe. Reported — the cBioPortal-backed AI-HOPE conversational-agent family documents the assembly class (AI-HOPE-WNT, Front. Artif. Intell. 2025; AI-HOPE-TP53, Cancers 2025).
  • Run a GWAS on case-control genotype data (added 2026-06-14) — rung-2 PLINK2 skill recipe: sample QC (call rate, sex-check, het outliers) → variant QC (--geno/--maf/--hwe in controls) → LD prune → --pca 10 → PCA-adjusted --glm logistic regression → lambda_GC check, then hand genome-wide-significant loci to the GWAS Catalog skill for prior-art annotation. Translational Medicine focus-day recipe; cookbook’s first GWAS recipe. Proposed — grounded in Chang et al., GigaScience 4:7 (2015) and Marees et al., Int. J. Methods Psychiatr. Res. 27:e1608 (2018). Biobank-scale mixed-model path deferred — no regenie/SAIGE component catalogued (see Missing components).
  • Build a pharmacogenomic dosing report from a patient’s diplotypes (added 2026-06-14) — rung-2 ClinPGx skill recipe: diplotype→metabolizer-phenotype translation via the CPIC PostgREST API → per-drug CPIC/DPWG dosing lookup → cited drug gene phenotype recommendation table, with explicit “no actionable guidance” flagging and a DDInter phenoconversion overlay noted. Translational Medicine focus-day recipe; cookbook’s first PGx-dosing recipe, kept distinct from the germline-pathogenicity variant-interpretation recipe. Proposed — grounded in the CPIC guideline corpus (Amstutz et al., Clin. Pharmacol. Ther. 2018; Molden & Jukić, Front. Pharmacol. 2021).
  • Infer cell-cell communication from single-cell RNA-seq (added 2026-06-13) — rung-2 LIANA-MCP recipe: load an annotated AnnData → ls_ccc_method → multi-method communicate (CellPhoneDB/Connectome/NATMI/SingleCellSignalR) → rank_aggregate consensus ligand-receptor tetrads → circle_plot/ccc_dotplot, consuming the annotated object from the scRNA-seq QC recipe. Molecular and Cellular Biology focus-day recipe; cookbook’s first cell-cell-communication recipe. Proposed — no documented LLM-driven LIANA-MCP assembly; grounded in Dimitrov et al., Nat. Commun. 13:3735 (2022), a 2026 consensus-LIANA application (Wei et al., PLOS ONE 2026), and the method-disagreement benchmark (Xie et al., Biomolecules 13:1211 (2023)).

Flagged for review

None.

Deferred — next-run priority

  • Select the best experimental PDB structure for a target before docking — rung-2 fit for the PDB MCP server (search_by_uniprot → resolution / R-free / chain coverage / ligand-bound vs apo / alternate-conformation pre-flight). Considered for this Integrative Structural and Computational Biology focus day but deferred because the AlphaFold-triage recipe already covers the optional PDB cross-check at one step; promoting it to its own recipe needs a clearer use case (“you have a PDB ID and want to know whether it’s the right one”). Revisit on the next ISCB focus day.
  • Pseudobulk single-cell DE end-to-end — natural composition of the QC + Scanpy pseudobulk + PyDESeq2 path; deferred because the Scanpy-MCP pseudobulk aggregation flow is mentioned but not yet documented as a focused end-to-end workflow.
  • Process resting-state EEG for spectral features end-to-end — natural companion to the HRV recipe using the NeuroKit2 skill on EEG channels (bandpower, microstates, complexity measures). Defer to a future Neuroscience focus day; closest analogous LLM workflow is EEGAgent (Yan et al., AAAI-26).
  • Browse OpenNeuro for MRI/EEG/MEG datasets matching a task or modality — clean rung-2 fit for the OpenNeuro MCP; GraphQL-driven dataset discovery (task name, modality, subject count, snapshot inspection) as the imaging counterpart to the existing DANDI/NWB recipe. Considered for this Neuroscience focus day but deferred to keep within the soft cap; revisit on the next Neuroscience pass.
  • Query the Allen Brain Atlas for cell-type / connectivity / expression context — clean rung-2 fit for the allenbrain-mcp (Alpha) — RMA queries, mouse connectivity experiments, brain-structure ontologies, 3D expression grid download. Deferred because the upstream MCP is Alpha with no LICENSE file; revisit after upstream clarifies licensing.
  • Choose an integration method via scIB benchmarking — companion to the integration recipe; needs a documented scib-metrics driver in the catalog before composing.
  • Prepare an AlphaFold-Multimer model of a protein-protein complex for interface analysis — natural companion to the AlphaFold triage recipe but blocked on co-folding / Multimer tooling not yet in catalog/tools/; revisit after the next Integrative Structural and Computational Biology pass.
  • RNA velocity / latent-time analysis with scVelo — natural rung-2 follow-on to the new Census atlas and integration recipes; deferred because most published scVelo case studies require upstream velocyto / STARsolo / kallisto-bustools spliced-counts pipelines that aren’t yet in the catalog (a CLI-only step before Claude takes over). Revisit after either the scVelo SKILL.md adds a worked example or a velocity-aware aligner skill is catalogued.
  • Drive pySCENIC end-to-end (motif filtering + AUCell) — natural rung-3 toolbelt extension of the new GRN-inference recipe; deferred because pySCENIC (the cisTarget + AUCell stack downstream of Arboreto) is not yet wrapped as a Claude skill or MCP server. Surfaced as a Missing component note.

Missing components

  • Biobank-scale mixed-model GWAS wrapper (regenie / SAIGE / BOLT-LMM) — the catalogued PLINK2 skill covers array/small-cohort fixed-effect GWAS (now the case-control GWAS recipe), but cohorts of 100k+ samples with relatedness/structure need a linear mixed model to control inflation. No regenie/SAIGE/BOLT-LMM component is Claude-installable today. Would unlock a biobank-scale sibling recipe. Surfaced 2026-06-14.
  • pySCENIC wrapper (cisTarget + AUCell) — no Claude-installable component for the steps downstream of GRNBoost2 (motif filtering against cisTarget databases, per-cell regulon AUCell scoring). Would unlock a full SCENIC-pipeline recipe on top of the new GRN-inference recipe. Surfaced 2026-05-28.
  • ADMET-AI / AdmetLab 3.0 / Deep-PK wrapper — none of the leading ML-based ADMET predictors (Swanson et al., Bioinformatics 2024; Fu et al., NAR 2024) has a Claude-installable wrapper (skill, MCP server, or plugin) in catalog/tools/ today. Their inclusion would upgrade the Estimate PK properties recipe from descriptor-and-analog-anchored estimation to a defensible ML-prediction layer for endpoints (CYP isoform IC50, hERG IC50, microsomal clearance) that pure descriptors miss. Surfaced 2026-05-27.
  • Co-folding / AlphaFold-Multimer / Boltz-2 wrapper — no Claude-installable component for protein-protein or protein-ligand co-folding currently in the catalog. Would unlock a complex-modelling companion to the AlphaFold triage recipe. Surfaced 2026-05-27.
  • MARRVEL-MCP (bioRxiv 2025-11-26, hyunhwan-bcm/MARRVEL_MCP) — rare-disease variant-interpretation MCP server with 39 tools and a published 95%-accuracy benchmark on 45 expert-curated tasks. Would strengthen the Interpret a clinical variant recipe from Proposed to Reported and open a Mendelian-disease-specific recipe. Surfaced 2026-05-23.

User requests (open)

  • [#41 @goodb 2026-06-15] queue: recipes feedback-on=run-functional-enrichment-on-a-gene-list sentiment=got stuck author=@goodb issue=41
  • [#43 @goodb 2026-06-15] queue: recipes report=composition outcome=worked problem_class=Knowledge synthesis author=@goodb issue=43

User requests (closed this run)

None.

Composition reports

Demand signal from the Composer plugin (/composer:compose). The responder routes report=composition entries into ## User requests (open) like any other request; process them there each run (promote a success, write a recipe for a gap), then keep a rolling tally here so the directed pass can prioritize the problem classes and subject areas scientists actually compose against. Keep the last ~15 lines.

Format: - YYYY-MM-DD outcome=<worked|gap|failed> problem_class=<…> → <what shipped / note>

None.