Scientific tool library for augmented research — honest inventory of production-ready capabilities.
Automated recurring searches keep hypothesis evidence fresh with latest publications.
Runs every 6 hours via systemd · API Status
The Forge provides computational tools that agents invoke during debates to strengthen arguments with evidence:
Each tool execution is logged for reproducibility and cost tracking.
Real Inventory: 147 tools currently available. This is our honest, working tool library — not aspirational vaporware.
Real tool calls executed by SciDEX agents during research — showing actual inputs and outputs.
Query protein-protein interaction network from STRING DB with confidence scores and evidence types.
gene_symbols=['PLA2G6', 'GPX4', 'PRDX6', 'FTH1', 'DMT1', 'SLC11A2'], score_threshold=400
Search PubMed for scientific papers. Returns PMIDs, titles, authors, abstracts.
query=MFN2 mitochondrial dynamics tau Alzheimer disease fusion fission, max_results=10
Query Allen Brain Atlas for brain region-specific gene expression data.
gene_symbol=CLU, max_results=25
Multi-step workflows that chain tools together — output of one feeds into the next. 5 chains, 0 total executions.
API: /api/forge/chains
Comprehensive gene research: gene info -> protein interactions -> disease associations -> clinical trials
Collect evidence for a hypothesis: PubMed search -> Semantic Scholar -> pathway analysis
Full protein characterization: UniProt annotation -> AlphaFold structure -> STRING interactions -> drug targets
Assess clinical significance of a gene variant: ClinVar -> gnomAD -> GWAS -> disease associations
Brain expression analysis: Allen Brain expression -> cell types -> GTEx tissue expression -> brain region context
Data endpoint: /api/forge/analytics
Tool calls by hour (UTC) — success / errors
Actual tool call volume grouped by tool type
Search PubMed for scientific papers. Returns PMIDs, titles, authors, abstracts.
Search ClinicalTrials.gov for trials: NCT ID, status, phase, conditions, interventions.
Fetch full abstract for a PubMed article by PMID.
Get gene annotation from MyGene.info: symbol, name, summary, aliases, GO terms.
Search Semantic Scholar for papers with citation counts and abstracts.
Convenience function combining PubMed, Semantic Scholar, and trials for comprehensive topic research.
Query protein-protein interaction network from STRING DB with confidence scores and evidence types.
Query Reactome pathway database for biological pathways involving a gene.
Retrieve clinical genetic variants from ClinVar with clinical significance and review status.
Get comprehensive protein annotation from UniProt: function, domains, PTMs, subcellular location.
Query Allen Brain Atlas for brain region-specific gene expression data.
Get disease associations for a gene from Open Targets platform.
Get genes associated with a disease from DisGeNET with association scores.
Query Human Protein Atlas for tissue and cell-specific expression and subcellular localization.
Query Allen Brain Cell Atlas for cell-type specific gene expression (SEA-AD, ABC Atlas)
Get disease associations for a gene from DisGeNET with scores and supporting evidence.
Search OpenAlex (250M+ works) for scholarly articles with citation counts, topics, open access status. Broader coverage than PubMed with better bibliometrics.
Functional enrichment analysis for gene lists via STRING DB (GO terms, pathways, diseases).
Run pathway/GO enrichment analysis on gene lists via Enrichr API.
Query NHGRI-EBI GWAS Catalog for genetic associations with traits or genes.
Query KEGG pathway database for biological pathways involving a gene with pathway diagrams.
Find drugs targeting a specific gene/protein from ChEMBL database with activity data.
Get gene expression levels across tissues from GTEx portal.
Get disease annotation from MyDisease.info: ontology, CTD data.
Get comprehensive gene annotation from Ensembl: coordinates, biotype, cross-references, mouse orthologs
Query OMIM for genetic diseases and phenotypes associated with a gene.
Automated pipeline that searches PubMed for new papers related to top hypotheses and updates evidence
Fetch AlphaFold protein structure predictions with confidence scores and 3D coordinates.
Query DGIdb for drug-gene interactions and druggability categories. Aggregates DrugBank, PharmGKB, TTD, ChEMBL, and clinical guideline data.
Search Europe PMC for biomedical literature with MeSH terms and citation counts
Query DrugBank for comprehensive drug information including targets, indications, and pharmacology.
Search NCBI GEO for gene expression and genomics datasets.
Get developmental brain gene expression from BrainSpan Atlas across human brain development (8 weeks post-conception to 40 years).
Query CZI CellxGene Discovery API for single-cell datasets. Returns brain/neurodegeneration datasets with cell type annotations, tissue info, and disease context.
Query EBI Proteins API for disease-associated protein variants with clinical annotations
Search Human Phenotype Ontology (HPO) for phenotype terms and gene/disease associations. Standardizes clinical phenotype vocabulary for neurodegeneration research.
Query InterPro for protein domain and family annotations (Pfam, SMART, PANTHER)
Query Mouse Genome Informatics for mouse models, phenotypes, and gene homologs.
Search Pathway Commons for pathways across Reactome, KEGG, PANTHER, NCI-PID
Query PubChem compound database for molecular structures and properties.
Query STITCH database for chemical-protein interactions. Aggregates data from experiments, databases, text mining, and predictions.
Query gnomAD for population variant frequency and gene constraint metrics (pLI, o/e ratios). Returns LoF/missense variants with allele frequencies and ClinVar pathogenic counts.
Query COSMIC (via Open Targets) for somatic cancer mutations. Returns cancer associations with somatic mutation evidence scores from Cancer Gene Census and IntOGen.
Query ClinGen for curated gene-disease validity classifications. Expert panels rate evidence from Definitive to Refuted. Gold-standard resource for determining whether a gene causes a specific disease.
Query EBI IntAct for experimentally validated molecular interactions. Aggregates data from BioGRID, MINT, DIP. Returns interactors with detection methods and publication evidence.
Query JASPAR database for transcription factor binding motifs and regulatory elements.
Query PharmGKB for pharmacogenomics drug-gene relationships. Returns clinical annotations linking genetic variants to drug response (efficacy, toxicity, dosage), level of evidence (1A–4), and CPIC guidelines.
Query EBI QuickGO for Gene Ontology annotations — biological process, molecular function, and cellular component terms for a gene.
Query Agora (AMP-AD Knowledge Portal, Sage Bionetworks) for Alzheimer's Disease multi-omic gene target scoring. Integrates RNA-seq, proteomics, metabolomics, network centrality, and genetic risk evidence across AD brain datasets. Returns nomination status, expression changes in AD brain (RNA + protein), genetic risk association (IGAP), and evidence across modalities. Essential for AD target prioritization.
Query Bgee for gene expression across anatomical structures and developmental stages. Integrates RNA-seq data with developmental stage and cell type context — more granular than GTEx for brain regions.
Query BioGRID database for experimentally verified protein-protein interactions with evidence types.
Search EBI BioStudies and ArrayExpress for transcriptomics, proteomics, and functional genomics datasets. Returns accession numbers, organism, study type. Complements NCBI GEO with European datasets and ArrayExpress studies.
Query EBI Complex Portal for experimentally validated protein complexes. Returns complex membership, subunit lists, and molecular functions. Essential for mechanistic interpretation of disease-associated proteins.
Search EBI Ontology Lookup Service (OLS4) across 300+ biomedical ontologies for disease, phenotype, molecular function, or chemical terms. Resolves free-text names to canonical ontology IDs: HPO (HP:), Disease Ontology (DOID:), MONDO, EFO, Gene Ontology (GO:), ChEBI. Essential for analyses that need standard identifiers for diseases or biological processes. Supports filtering by ontology and exact-match queries. Returns IDs, labels, descriptions, and synonyms.
Search ENCODE for epigenomics experiments (ChIP-seq, ATAC-seq, Hi-C) targeting a gene. Returns released experiments with biosample, assay type, and accession IDs. Critical for interpreting non-coding GWAS variants by finding TF binding sites and chromatin accessibility near neurodegeneration risk loci.
Query Ensembl Regulatory Build for regulatory elements in the genomic neighborhood of a gene. Returns promoters, enhancers, CTCF binding sites, and open chromatin regions derived from ENCODE/Roadmap data. Critical for interpreting non-coding GWAS variants near neurodegeneration loci (BIN1, CLU, PICALM, APOE). Reveals the regulatory landscape that controls gene expression.
Query EMBL-EBI Expression Atlas for differential expression experiments. Returns experiments where a gene is differentially expressed across conditions, tissues, and diseases.
Query GTEx v8 for cis-eQTLs in brain tissues: genetic variants that regulate gene expression in frontal cortex, hippocampus, substantia nigra and 6 other brain regions. Connects GWAS hits to causal gene regulation.
Query IMPC (International Mouse Phenotyping Consortium) for statistically significant phenotypes observed in knockout mice (p<0.0001). Covers 20+ biological systems including neurological, cardiovascular, metabolic, and immune phenotypes. Provides direct in vivo evidence of gene function — essential for validating disease hypotheses about neurodegeneration genes like TREM2, GBA, LRRK2, and PSEN1.
Query JensenLab DISEASES (STRING group) for text-mining gene-disease confidence scores derived from 500M+ MEDLINE abstracts. Provides independent evidence scores complementary to DisGeNET and ClinGen. Covers rare and common diseases with automated mining of scientific literature. High-confidence associations reflect strong co-mention signals across publications.
Query MSigDB gene set membership for a gene via Enrichr genemap. Returns which Hallmark, KEGG, Reactome, WikiPathways, and GO gene sets directly contain the query gene. Complements enrichr_analyze (which runs enrichment on a gene list) by answering 'which named pathways include this gene?' — useful for contextualizing a single candidate gene.
Query Monarch Initiative for disease-gene-phenotype associations. Integrates OMIM, ClinVar, HPO, MGI, ZFIN data. Supports disease→gene and gene→disease queries.
Fetch official NCBI gene summary, aliases, chromosomal location, and MIM IDs via the NCBI Datasets v2 API. Returns a concise functional overview from RefSeq curators — useful as a first-pass description of gene function before querying pathway or disease databases. Covers all human genes including rare neurodegeneration-associated genes.
Search NIH RePORTER for funded research grants on a topic. Returns project titles, PIs, institutions, award amounts, and abstract excerpts. Useful for mapping the funding landscape for neurodegeneration research areas, identifying active investigators, and understanding NIH research priorities. Covers all NIH institutes including NIA, NINDS, and NIMH.
Query OmniPath for directed signaling interactions and post-translational modifications (PTMs). Integrates 100+ databases (BioGRID, HPRD, PhosphoSitePlus, SignaLink, Reactome, CellTalkDB) with stimulation/inhibition directionality. Unlike STRING, focuses on regulatory direction — which kinase activates which receptor. Includes ligand-receptor interactions for cell-cell communication.
Query Open Targets Platform for drugs targeting a gene, with clinical phases, mechanisms of action, and disease indications. Focuses on clinical evidence and approved drugs — complements ChEMBL bioactivity data for drug repurposing.
Query Open Targets Genetics for GWAS loci linked to a gene via Locus-to-Gene (L2G) ML scoring. Maps genetic variants to probable causal genes using eQTLs, chromatin accessibility, and functional genomics. Essential for interpreting AD/PD GWAS hits.
Query Open Targets for baseline RNA expression across 100+ tissues for a gene, aggregated from GTEx and Expression Atlas. Returns TPM-like values and z-scores per tissue. Includes a brain_only flag to filter to CNS/neuronal tissues, making it directly useful for neurodegeneration context: microglia (TREM2), neurons (SNCA/MAPT), oligodendrocytes. Complements gtex_eqtl (genetic effects) and bgee_expression (developmental) with baseline expression profiles.
Query Open Targets Platform for drug tractability and modality assessments. Returns tractability buckets for small molecules (clinical/preclinical precedent, structural features), antibodies (membrane protein evidence), and other modalities. Critical for prioritizing therapeutic targets — especially relevant for neurodegeneration targets like LRRK2 (kinase), TREM2 (receptor), and GBA (enzyme).
Search RCSB Protein Data Bank for experimental protein structures (X-ray, cryo-EM, NMR). Complements AlphaFold predictions with experimentally validated structural data including drug binding sites.
Get canonical cell type marker genes from PanglaoDB scRNA-seq database. Covers microglia, astrocytes, neurons, OPCs, DAM, oligodendrocytes, endothelial, pericytes.
Query NIH Pharos TCRD for drug target development level (TDL): Tclin (approved drugs), Tchem (bioactive molecules), Tbio (biological knowledge only), Tdark (poorly characterized). Returns target family, disease associations, drugs and bioactive compounds. Essential for drug target prioritization in neurodegeneration.
Query WikiPathways community pathway database for biological pathways containing a gene. Complements Reactome and KEGG with community-curated pathways including disease-specific and rare pathway annotations. Returns pathway IDs, names, species, and visualization URLs.
Query GTEx v8 for splicing QTLs (sQTLs) in brain: genetic variants that alter RNA splicing patterns across 9 brain regions. Complementary to eQTLs — many disease risk variants act by changing splice junction usage rather than expression level. Key examples: MAPT H1/H2 haplotype sQTLs (tau isoform switching), BIN1 hippocampal sQTLs. Returns per-tissue splice junction associations with variant IDs.
Query HUGO Gene Nomenclature Committee (HGNC) for authoritative gene names, aliases, previous symbols, gene family membership, locus type (protein-coding/lncRNA/pseudogene), chromosomal location, and cross-references to Ensembl/UniProt/OMIM/RefSeq/MGI. Gold-standard for gene symbol disambiguation and family classification.
Query Harmonizome for gene-associated datasets across 114 gene-set libraries (KEGG, Reactome, OMIM, GO, GTEx, GWAS, ChEMBL, etc.). Enables rapid cross-database characterization of any gene.
Query Open Targets for mouse model phenotypes associated with a gene. Returns experimentally observed phenotypes from knockout/transgenic mouse models aggregated from IMPC, MGI, and other sources. Provides phenotype class summaries (neurological, behavioral, metabolic, etc.) and biological model details (allelic composition, genetic background, literature). Useful for assessing whether modifying a gene has neurological or disease-relevant consequences in preclinical models.
Extract figures from a scientific paper by PMID. Returns figure captions, image URLs, and descriptions via PMC BioC API, Europe PMC full-text XML, or open-access PDF extraction. Use when you need to see visual evidence (pathway diagrams, heatmaps, microscopy) from a cited paper.
Retrieve UniProt/Swiss-Prot curated protein feature annotations for a gene: post-translational modifications (phosphorylation, ubiquitination, acetylation, methylation), active sites, binding sites, glycosylation, disulfide bonds, signal peptides, and natural variants. Especially valuable for neurodegeneration research: tau (MAPT) phosphorylation landscape, alpha-synuclein (SNCA) PTMs, APP cleavage sites. Returns structured feature list with positions and evidence counts.
Annotate genetic variants using Ensembl Variant Effect Predictor (VEP). Accepts dbSNP rsIDs or HGVS notation and returns predicted molecular consequences (missense, splice, frameshift), SIFT/PolyPhen-2 pathogenicity scores, amino acid changes, and impact classification (HIGH/MODERATE/LOW/MODIFIER). Essential for interpreting neurodegeneration GWAS variants: APOE4=rs429358, LRRK2 G2019S=rs34637584, GBA N370S=rs76763715.
Get articles that cite a specific paper via Europe PMC
Query ProteomicsDB (TUM/Kuster lab) for mass spectrometry-based protein abundance across human tissues. Measures actual protein levels (iBAQ normalized intensity) across 60+ tissues/fluids. Complements RNA expression atlases (GTEx, HPA) with protein-level data — critical because mRNA and protein levels often diverge. Accepts gene symbol or UniProt accession.
Query Alliance of Genome Resources (AGR) for cross-species gene orthologs using 8 integrated algorithms (OrthoFinder, PANTHER, Ensembl Compara, OMA, InParanoid, Phylome, OrthoMCL, HGNC). Returns orthologs in mouse, rat, zebrafish, fly, worm, and yeast with prediction method count and best-score flag. Critical for identifying model organism experiments that validate neurodegeneration mechanisms.
Query Allen Brain Atlas for ISH expression data across brain regions.
Query BindingDB for protein-ligand binding affinity measurements (Ki, Kd, IC50, EC50) from the primary literature. Accepts gene symbol or UniProt accession. Returns compound names, affinities, SMILES, PubMed IDs, and BindingDB URLs. Supports optional max_ic50_nm filter for potent binders. Essential for drug target prioritization (BACE1, LRRK2, GSK3B, CDK5).
Get CIViC (Clinical Interpretation of Variants in Cancer) expert-curated clinical variant interpretations for a gene. Returns variant names, HGVS expressions, variant types, and CIViC URLs. Useful for variant effect classification, functional evidence in overlapping cancer/neurodegeneration genes (IDH1, PTEN, ATM, BRCA2, TP53), and mechanistic evidence for rare disease variant interpretation.
Query CZ CELLxGENE Discover 'Where is My Gene' (WMG v2) API for quantitative single-cell gene expression across human cell types. Returns mean log2(CPM+1) expression, percent expressing cells, and total cell count per cell-type per tissue. Brain-relevant cell types (microglia, neurons, astrocytes, oligodendrocytes, pericytes) are prioritized. Distinct from cellxgene_gene_expression which only lists datasets. Critical for neurodegeneration: understanding which brain cell types express TREM2/APOE/APP/SNCA/MAPT at the single-cell level.
Search ChEMBL (EBI) for small molecules, drugs, and chemical probes by name. Returns structural information, clinical development phase, ATC pharmacological classification, molecular properties (MW, LogP, HBD/HBA), and InChI key. Compound-centric complement to chembl_drug_targets (gene-centric). Useful for looking up a drug candidate's ChEMBL ID and clinical phase, or finding related compound series.
Drug compounds and bioactivity data for a gene target from the ChEMBL database of bioactive molecules.
Query ClinGen for curated gene-disease validity classifications.
Fetch clinical genetic variants from NCBI ClinVar. Returns pathogenicity, review status, and associated conditions.
Search ClinicalTrials.gov for clinical trials related to genes, diseases, or interventions. Returns NCT IDs, status, phase, conditions, interventions, enrollment, and sponsor info.
Retrieve publication metadata from CrossRef by DOI. Returns title, authors, journal, year, citation count, publisher, open access PDF link, subject areas, and funders. CrossRef indexes 150M+ scholarly works across all publishers — covers papers from any journal (not just biomedical), including clinical trials, meta-analyses, and specialized neuro-chemistry publications not in PubMed. Essential for DOI-based paper lookup when PMID is unavailable.
Search for bioRxiv/medRxiv preprints via CrossRef API. Returns preprints indexed by CrossRef with title, authors, DOI, posted date, and abstract snippet. Useful for finding the latest findings before peer review, especially for fast-moving topics like tau pathology, neuroinflammation, TREM2 biology, Parkinson's alpha-synuclein research, and ALS/FTD mechanisms.
Query DGIdb for drug-gene interactions and druggability categories. Aggregates DrugBank, PharmGKB, TTD, ChEMBL, and clinical guideline data.
Find diseases related to a query disease based on shared gene-disease associations in DisGeNET. Uses top-scoring genes for the query disease to identify comorbid or mechanistically related conditions. Helps map shared molecular substrates across neurodegeneration (e.g. diseases sharing APOE, LRRK2, or GBA pathways). Requires DISGENET_API_KEY for full results.
Get genes associated with a disease from DisGeNET with association scores.
Get disease associations for a gene from DisGeNET with scores and supporting PMIDs.
Query EBI eQTL Catalog v2 for tissue/cell-type-specific expression QTLs. Covers 100+ studies beyond GTEx including microglia, iPSC-derived neurons, brain organoids, and disease cohorts. Returns rsid, variant, p-value, beta, SE, MAF, tissue_label, and qtl_group. Supports optional tissue_keyword filter (e.g. 'microglia', 'brain', 'iPSC').
Gene set enrichment against GO Biological Process. Enter a gene list to find enriched pathways.
Retrieve multi-source phenotype and disease associations for a gene via Ensembl REST API. Aggregates disease-gene links from Orphanet, OMIM/MIM morbid, NHGRI-EBI GWAS Catalog, DECIPHER, ClinVar, and UniProtKB into a unified view with MONDO/HP/Orphanet ontology accessions and data source provenance. Distinct from omim_gene_phenotypes (OMIM only), clinvar_variants (variant-level), and monarch_disease_genes (Monarch KG). Useful for comprehensive disease context, rare disease associations via Orphanet, and multi-source phenotype validation.
Query FinnGen R10 (N=520,000 Finns, 2,408 endpoints) for fine-mapped genetic loci. Returns SuSiE credible sets with lead SNP, gene, p-value, beta, and cross-trait annotations.
Genome-wide association study hits from the NHGRI-EBI GWAS Catalog. Query by gene or trait.
Query EBI GWAS Catalog for all phenotypic associations reported for a specific genetic variant (rsID). Returns distinct traits grouped by study count and best p-value. Complements gwas_genetic_associations (gene-centric): this tool lets you see the full phenotypic spectrum of a known variant. Essential for major neurodegeneration variants: APOE4=rs429358 (Alzheimer, lipids, cognition), LRRK2 G2019S=rs34637584 (Parkinson), GBA N370S=rs76763715.
Look up any human gene — returns full name, summary, aliases, and gene type from MyGene.info.
Query Genomics England PanelApp for disease gene panel memberships. Used by NHS Genomic Medicine Service for rare disease diagnosis. Returns panel name, disease group, confidence level (green/amber/red), mode of inheritance, penetrance, phenotypes, and gene evidence. Covers hereditary dementias, Parkinson, motor neuron disease, ataxias, and leukodystrophies.
Protein expression across human tissues and cell types from the Human Protein Atlas. Includes subcellular localisation.
Query the KEGG Disease database for curated disease entries and their causal/associated genes. Returns KEGG disease IDs, gene lists with subtypes (e.g. AD1/APP, AD17/TREM2), approved drugs, and linked pathways. Disease-centric view distinct from kegg_pathways (gene-to-pathway). Valuable for getting the official KEGG gene list for neurodegeneration diseases with clinical subtypes.
Search LIPID MAPS for lipid structures, classifications, and biological roles via KEGG cross-reference. Returns LMID, formula, exact mass, main/sub class, InChIKey, HMDB/ChEBI IDs, and LIPID MAPS URL. Covers glycerophospholipids, sphingolipids, gangliosides, ceramides, oxysterols, and plasmalogens — all implicated in neurodegeneration (NPC, Alzheimer, Parkinson, ALS).
Search the NIH Metabolomics Workbench for public metabolomics datasets. Returns study IDs, titles, species, analysis type (LC-MS, GC-MS, NMR), sample counts, and URLs. Useful for finding published CSF, brain region, or plasma metabolomics datasets for neurodegeneration diseases. Complements RNA/protein atlases with metabolite-level evidence.
Get age-related methylation changes
Query cross-species methylation conservation
Get CpG island information for gene regions
Query developmental stage methylation dynamics
Search differential methylation between conditions
Search disease-associated methylation changes
Query DNA methylation studies for specific genes
Compare methylation patterns across tissues
Get tissue-specific methylation patterns
Query Monarch Initiative for disease-gene-phenotype associations from OMIM, ClinVar, HPO
Retrieve NCBI Gene Reference Into Function (GeneRIF) linked publications for a gene. GeneRIF is a curated set of brief functional annotations each backed by a PubMed paper. Returns titles, journals, and URLs for the top-cited papers supporting gene function annotations — a fast path to high-quality mechanistic evidence for any neurodegeneration gene.
Look up official NCBI Medical Subject Headings (MeSH) descriptors for diseases, pathways, and biological concepts. Returns the preferred descriptor name, hierarchical tree codes (e.g. C10.228.140.380 for Alzheimer Disease), scope note definition, and entry terms (all synonyms accepted by PubMed). Essential for standardizing terminology, discovering correct PubMed search terms, and navigating the MeSH vocabulary hierarchy from broad to specific.
Search the NCBI Sequence Read Archive (SRA) for public high-throughput sequencing datasets. Returns run accessions, titles, organisms, library strategy (RNA-Seq, ATAC-seq, ChIP-seq, scRNA-seq), and center names. Essential for finding publicly available neurodegeneration patient or model datasets for reanalysis.
Look up a variant in NCBI dbSNP for chromosomal position (GRCh38), reference/alternate alleles, functional class, gene context, and population allele frequencies from studies including 1000 Genomes, ALFA, and TOPMED. Complements Ensembl VEP (consequence prediction) and gnomAD (constraint) with official dbSNP registration metadata. Essential for characterizing a specific variant rsID.
Query OmniPath for post-translational modification (PTM) interactions integrating PhosphoSitePlus, PhosphoELM, SIGNOR, DEPOD, dbPTM, HPRD and 30+ other sources. Returns enzymes modifying the query gene AND substrates the gene modifies — phosphorylation, ubiquitination, acetylation, SUMOylation, methylation. Critical for tau kinase/phosphatase networks, alpha-synuclein ubiquitination (PINK1/Parkin), LRRK2 substrates, TDP-43/FUS in ALS.
Get top-scored target genes for a disease from Open Targets Platform, integrating 14+ evidence types (GWAS, rare genetics, expression, pathways, animal models, literature). Disease-centric: given a disease name, returns ranked gene list with overall association score and per-evidence-type breakdown. Essential for target prioritization debates. Distinct from open_targets_associations (gene→diseases).
Disease associations and therapeutic evidence for a gene from Open Targets Platform, scored across multiple evidence sources.
Query Open Targets for target safety liability evidence curated from literature, clinical genetics, and animal toxicology. Returns safety events, affected tissues, inhibition effects, and supporting PMIDs. Critical for neurodegeneration drug discovery: CNS off-target effects, gene essentiality, tissue-specific toxicity, and high-risk indications for targets like LRRK2, BACE1, MAPT, CDK5, GSK3B.
Query the MRC IEU OpenGWAS platform for phenome-wide association study (PheWAS) data. Returns all GWAS traits where a given variant reaches genome-wide significance across 10K+ studies including UK Biobank, FinnGen, and many GWAS consortia. Complements GWAS Catalog by covering unpublished and OpenGWAS-native summary statistics. Useful for characterizing pleiotropic variants like rs429358 (APOE4).
Search PGS Catalog (EMBL-EBI) for published polygenic risk score (PRS) models for a disease. Returns multi-SNP scoring models with variant counts, effect weight methods, publication DOI/year, and FTP download links. Covers 47+ Alzheimer disease PRS, 11+ Parkinson disease PRS, and hundreds of cognitive/brain trait models. Complements GWAS tools (single variants) with complete polygenic models ready for individual risk stratification. Essential for precision medicine analyses in neurodegeneration.
Get canonical cell type marker genes from PanglaoDB scRNA-seq database. Covers microglia, astrocytes, neurons, OPCs, DAM, oligodendrocytes.
Ingest a list of paper dicts into the local PaperCorpus cache for persistent storage. Each paper needs at least one ID (pmid, doi, or paper_id).
Search across PubMed, Semantic Scholar, OpenAlex, and CrossRef with unified results and local caching. Use providers param to filter to specific sources.
Start a stateful multi-page search session. Call again with incremented page param to fetch subsequent pages.
Search Paperclip MCP for biomedical papers (8M+ across arXiv, bioRxiv, PMC, OpenAlex, OSF). Uses hybrid BM25+embedding search with TL;DR summaries.
Query PharmGKB for pharmacogenomics drug-gene relationships. Returns clinical annotations linking genetic variants to drug response.
Find bioassay-confirmed active compounds against a protein target in PubChem BioAssay. Returns assay IDs linked to the gene and CIDs of active compounds. Useful for identifying chemical probes, drug leads, validated inhibitors, and tool compounds for neurodegeneration target validation (BACE1, CDK5, GSK3B, LRRK2, GBA, HDAC6, PARP1).
Search PubMed for papers by keyword. Returns titles, authors, journals, PMIDs.
Extract standardized gene, disease, chemical, and variant mentions from PubMed literature using NCBI PubTator3 AI annotation. Resolves synonyms to canonical IDs (NCBI Gene IDs, MeSH disease IDs). Useful for finding papers where specific genes and diseases co-occur, building evidence chains for hypotheses, and disambiguating biomedical entity names.
Search Reactome for pathways by concept name and return constituent genes for each pathway. Complements the gene→pathway direction by going pathway-name→genes. Essential for building mechanistic gene sets for neurodegeneration analyses: mitophagy, tau clearance, NLRP3 inflammasome, amyloid processing, autophagy, endosomal sorting, complement activation, mTOR signaling.
Look up biological pathways a gene participates in, from Reactome.
Query STRING DB for a multi-gene functional interaction network with per-channel evidence scores: coexpression (escore), experimental binding/co-IP (ascore), text-mining (tscore), curated database (dscore), neighborhood (nscore), and gene fusion (fscore). Unlike the basic STRING protein interactions tool, this returns the complete evidence breakdown for every pairwise link so you can distinguish mechanistic (experimental) from correlative (coexpression/text-mining) evidence. Essential for assessing gene module cohesion in AD (TREM2/TYROBP/SYK/PLCG2), PD (PINK1/PARK2/LRRK2/SNCA), and ALS/FTD (TDP-43/FUS/C9orf72) networks.
Find physical protein-protein interactions from the STRING database. Enter 2+ gene symbols.
Cross-reference a compound across 40+ chemistry and pharmacology databases via EBI UniChem. Given a compound name or InChI key, returns the compound's IDs in ChEMBL, DrugBank, PubChem, BindingDB, DrugCentral, KEGG, ChEBI, and more. Essential for integrating multi-database drug evidence and resolving compound identity across sources.
Comprehensive protein annotation from UniProt/Swiss-Prot: function, domains, subcellular location, disease associations.
Functional enrichment analysis via g:Profiler (ELIXIR bioinformatics platform). Tests a gene list against GO Biological Process, GO Molecular Function, GO Cellular Component, KEGG, Reactome, WikiPathways, Human Phenotype Ontology, and TRANSFAC databases. Complements Enrichr with different statistical correction (g:SCS method), independent database versions, and broader ontology coverage including HP phenotypes and miRTarBase targets.
Query FDA Adverse Event Reporting System (FAERS) via openFDA API for drug safety signals. Returns ranked adverse reactions with report counts and percentages from post-market surveillance. Useful for identifying neurological adverse events, CNS toxicity, ARIA risk, and safety liabilities for drugs being repurposed in neurodegeneration (donepezil, memantine, lecanemab, aducanumab, levodopa, riluzole).