Skip to content

NEWS.md

Latest commit

 

History

History
96 lines (71 loc) · 4.67 KB

NEWS.md

File metadata and controls

96 lines (71 loc) · 4.67 KB

TCMDATA 0.1.0

Released: 2026-04-01

This is the first public release of TCMDATA, a comprehensive R toolkit for Traditional Chinese Medicine (TCM) network pharmacology research.

🌿 Data Retrieval

  • search_herb(): Query herb–compound–target associations by herb name (Chinese, Pinyin, or Latin)
  • search_target(): Reverse lookup—find herbs targeting specific genes
  • search_disease(): Query DisGeNET (via DOSE) for disease-associated genes; supports UMLS CUI, exact name, and fuzzy matching
  • search_gene_disease(): Reverse lookup—find all diseases associated with given gene symbols or Entrez IDs
  • search_geo_datasets(): Query NCBI GEO for expression datasets (GDS/GSE) by disease keyword; filters by organism and dataset type

🔬 Molecule Detection (PubChem Integration)

  • resolve_cid(): Resolve compound names to PubChem CIDs
  • getprops(): Retrieve compound properties (MW, LogP, TPSA, etc.)
  • compound_similarity(): Tanimoto similarity search against PubChem
  • download_ligand_structure(): Download molecular structures (SDF/MOL2/PDB)
  • convert_structure(): Convert between molecular file formats

📊 Network Analysis

  • get_ppi(): Retrieve STRING PPI networks (wrapper around clusterProfiler::getPPI())
  • prepare_herb_graph(): Build herb–compound–target networks
  • ppi_subset(): Filter PPI networks by confidence score
  • compute_nodeinfo(): Calculate 17+ centrality metrics (degree, betweenness, closeness, MCC, DMNC, BN, EPC, radiality, Stress, etc.)
  • rank_ppi_nodes(): Rank nodes by multiple topological features
  • ppi_knock(): Evaluate network robustness via drug-attack simulation

🧬 Clustering Algorithms

  • run_louvain(): Louvain community detection
  • run_MCL(): Markov Clustering (MCL)
  • run_mcode(): MCODE algorithm for dense subgraph detection

📈 Enrichment Analysis

  • herb_enricher(): Herb-based over-representation analysis (ORA)
  • Compatible with GO/KEGG via clusterProfiler

🤖 Machine Learning Screening

  • run_ml_screening(): Run 6 ML algorithms with 3 validation modes

    • LASSO, Elastic Net, Ridge

    • Random Forest + Boruta

    • SVM-RFE

    • XGBoost

  • get_ml_consensus(): Consensus feature selection across methods
  • plot_ml_roc(), plot_ml_venn(): Visualize ML results

💡 AI-Powered Interpretation

  • tcm_setup(): Configure LLM provider (OpenAI, Anthropic, Gemini, Deepseek, aihubmix, and 10+ others)
  • tcm_interpret(): Interpret enrichment results, PPI data, or free text
  • draft_result_paragraph(): Generate manuscript-ready result paragraphs
  • tcm_interpret_schema(): Structured output with custom schemas
  • create_tcm_agent(), make_tcm_function(): Build reusable AI functions

🧠 AI Agent System (NEW)

  • tcm_agent(): One-shot natural language entry point with automatic task routing to appropriate tools
  • tcm_chat(): Interactive multi-turn chat session with built-in commands (/help, /artifacts, /history, /model, /stats, /clear, /quit)
  • route_tcm_task(): Rule-based task router supporting 11 task types (herb_lookup, disease_lookup, enrichment, ppi_analysis, ml_screening, pubmed, geo_search, visualization, interpretation, etc.) with Chinese and English keyword patterns
  • create_tcm_task_agent(): Create a configured aisdk agent with tools and skills for custom workflows
  • create_tcm_tools(): Generate 36 aisdk Tool objects wrapping TCMDATA functions; supports filtering by task_type or tool_names
  • tcm_init_skills(): Copy bundled skills to local directory for customization
  • tcm_use_skills() / tcm_reset_skills() / tcm_skill_dir(): Skill directory management

📚 Literature Mining

  • get_pubmed_data(): Search PubMed for TCM–disease literature
  • get_pubmed_table(): Export results as publication tables

📊 Visualization

  • tcm_sankey(): Herb–compound–target Sankey diagrams
  • ggdot_sankey(): Combined dot-Sankey plots for enrichment
  • ggdock(): Molecular docking affinity heatmaps
  • gglollipop(): Lollipop plots for enrichment results
  • plot_node_heatmap(): PPI centrality heatmaps
  • radar_plot(): Radar charts for multi-dimensional comparisons
  • go_barplot(), gocircle_plot(): GO enrichment visualizations
  • ggvenn_plot(): Venn diagrams
  • upsetplot(): UpSet plot for gene set intersection visualization (wrapper around aplotExtra::upset_plot())

📦 Bundled Datasets

  • gutMGene: Gut microbiota–metabolite associations
  • tf_targets: Transcription factor–target regulation pairs
  • dn_gcds, dn_otp: Diabetic nephropathy reference datasets (GeneCards & Open Targets)
  • deg_earlydn: DESeq2 results for early diabetic nephropathy (GSE142025)
  • demo_ppi: Example PPI network for tutorials
  • covid19: COVID-19 case study dataset