@bgicli/bgicli 2.1.1 → 2.2.0

package/data/skills/leads-literature-mining/SKILL.md

@@ -0,0 +1,68 @@
+<!--
+# COPYRIGHT NOTICE
+# This file is part of the "Universal Biomedical Skills" project.
+# Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>
+# All Rights Reserved.
+#
+# This code is proprietary and confidential.
+# Unauthorized copying of this file, via any medium is strictly prohibited.
+#
+# Provenance: Authenticated by MD BABU MIA
+
+-->
+
+---
+name: leads-literature-mining
+description: Review Automator
+keywords:
+  - literature-mining
+  - systematic-review
+  - meta-analysis
+  - pubmed
+  - evidence-synthesis
+measurable_outcome: Complete a systematic review screen of 100+ papers with >90% inclusion/exclusion accuracy compared to human baseline.
+license: CC-BY-4.0
+metadata:
+  author: Nature Communications 2025
+  version: "1.0.0"
+compatibility:
+  - system: Python 3.9+
+allowed-tools:
+  - run_shell_command
+  - web_fetch
+---
+
+# LEADS (Literature Mining Agent)
+
+A specialized LLM agent for automating systematic reviews and meta-analyses, capable of high-accuracy study selection and data extraction.
+
+## When to Use
+
+*   **Systematic Reviews**: Screening thousands of abstracts for inclusion criteria.
+*   **Data Extraction**: Pulling specific metrics (e.g., hazard ratios, sample sizes) from full-text PDFs.
+*   **Evidence Synthesis**: Aggregating findings across multiple studies.
+
+## Core Capabilities
+
+1.  **Study Selection**: Automated screening based on PICO criteria.
+2.  **Data Extraction**: Structured extraction of study characteristics and results.
+3.  **Quality Assessment**: Risk of bias evaluation.
+
+## Workflow
+
+1.  **Search**: Query PubMed/Embase.
+2.  **Screen**: Apply inclusion/exclusion criteria to abstracts.
+3.  **Extract**: Parse full text for data points.
+4.  **Report**: Generate PRISMA flow diagram and evidence table.
+
+## Example Usage
+
+**User**: "Perform a systematic review on the efficacy of CAR-T in solid tumors."
+
+**Agent Action**:
+```bash
+python -m leads.review --topic "CAR-T solid tumors" --criteria ./criteria.json
+```
+
+
+<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->

package/data/skills/ligandmpnn/SKILL.md

@@ -0,0 +1,170 @@
+---
+name: ligandmpnn
+description: >
+  Ligand-aware protein sequence design using LigandMPNN.
+  Use this skill when: (1) Designing sequences around small molecules,
+  (2) Enzyme active site design,
+  (3) Ligand binding pocket optimization,
+  (4) Metal coordination site design,
+  (5) Cofactor binding proteins.
+
+  For standard protein design, use proteinmpnn.
+  For solubility optimization, use solublempnn.
+license: MIT
+category: design-tools
+tags: [sequence-design, inverse-folding, ligand-aware]
+biomodals_script: modal_ligandmpnn.py
+---
+
+# LigandMPNN Ligand-Aware Design
+
+## Prerequisites
+
+| Requirement | Minimum | Recommended |
+|-------------|---------|-------------|
+| Python | 3.8+ | 3.10 |
+| CUDA | 11.0+ | 11.7+ |
+| GPU VRAM | 8GB | 16GB (T4) |
+| RAM | 8GB | 16GB |
+
+## How to run
+
+> **First time?** See [Installation Guide](../../docs/installation.md) to set up Modal and biomodals.
+
+### Option 1: Modal (recommended)
+```bash
+cd biomodals
+modal run modal_ligandmpnn.py \
+  --pdb-path protein_ligand.pdb \
+  --num-seq-per-target 16 \
+  --sampling-temp 0.1
+```
+
+**GPU**: T4 (16GB) | **Timeout**: 600s default
+
+### Option 2: Local installation
+```bash
+git clone https://github.com/dauparas/LigandMPNN.git
+cd LigandMPNN
+
+python run.py \
+  --pdb_path protein_ligand.pdb \
+  --out_folder output/ \
+  --num_seq_per_target 16
+```
+
+## Key parameters
+
+| Parameter | Default | Range | Description |
+|-----------|---------|-------|-------------|
+| `--pdb_path` | required | path | PDB with ligand |
+| `--num_seq_per_target` | 1 | 1-1000 | Sequences per structure |
+| `--sampling_temp` | "0.1" | "0.0001-1.0" | Temperature (string!) |
+| `--ligand_mpnn_use_side_chain_context` | true | bool | Use ligand context |
+
+## Ligand Specification
+
+### In PDB File
+Ligand must be present as HETATM records:
+```
+ATOM    ...protein atoms...
+HETATM  1  C1  LIG A 999      x.xxx  y.yyy  z.zzz  1.00  0.00           C
+```
+
+### Supported Ligand Types
+- Small molecules (HETATM)
+- Metals (Zn, Fe, Mg, Ca, etc.)
+- Cofactors (NAD, FAD, ATP)
+- DNA/RNA
+
+## Output format
+
+```
+output/
+├── seqs/
+│   └── protein.fa          # FASTA sequences
+└── protein_pdb/
+    └── protein_0001.pdb    # PDBs with designed sequence
+```
+
+## Sample output
+
+### Successful run
+```
+$ python run.py --pdb_path enzyme_substrate.pdb --out_folder output/ --num_seq_per_target 8
+Loading LigandMPNN model weights...
+Processing enzyme_substrate.pdb
+Found ligand: LIG (12 atoms)
+Generated 8 sequences in 3.1 seconds
+
+output/seqs/enzyme_substrate.fa:
+>enzyme_substrate_0001, score=1.45, global_score=1.38
+MKTAYIAKQRQISFVKSHFSRQLE...
+>enzyme_substrate_0002, score=1.52, global_score=1.41
+MKTAYIAKQRQISFVKSQFSRQLD...
+```
+
+**What good output looks like:**
+- Score: 1.0-2.0 (lower = more confident)
+- Ligand detected and incorporated in context
+- Active site residues preserved or optimized
+
+## Decision tree
+
+```
+Should I use LigandMPNN?
+│
+├─ What's in your binding site?
+│  ├─ Small molecule / ligand → LigandMPNN ✓
+│  ├─ Metal ion (Zn, Fe, etc.) → LigandMPNN ✓
+│  ├─ Cofactor (NAD, FAD, ATP) → LigandMPNN ✓
+│  ├─ DNA/RNA → LigandMPNN ✓
+│  └─ Nothing / protein only → Use ProteinMPNN
+│
+├─ What type of design?
+│  ├─ Enzyme active site → LigandMPNN ✓
+│  ├─ Metal binding site → LigandMPNN ✓
+│  ├─ Protein-protein binder → Use ProteinMPNN
+│  └─ De novo scaffold → Use ProteinMPNN
+│
+└─ Priority?
+   ├─ Solubility/expression → Consider SolubleMPNN
+   └─ Ligand context accuracy → LigandMPNN ✓
+```
+
+## Typical performance
+
+| Campaign Size | Time (T4) | Cost (Modal) | Notes |
+|---------------|-----------|--------------|-------|
+| 100 backbones × 8 seq | 15-20 min | ~$2 | Standard |
+| 500 backbones × 8 seq | 1-1.5h | ~$8 | Large campaign |
+
+**Throughput**: ~50-100 sequences/minute on T4 GPU.
+
+---
+
+## Verify
+
+```bash
+grep -c "^>" output/seqs/*.fa  # Should match backbone_count × num_seq_per_target
+```
+
+---
+
+## Troubleshooting
+
+**Ligand not recognized**: Check HETATM format, verify ligand residue name
+**Poor binding residues**: Increase sampling around active site
+**Missing contacts**: Verify ligand coordinates in PDB
+
+### Error interpretation
+
+| Error | Cause | Fix |
+|-------|-------|-----|
+| `RuntimeError: CUDA out of memory` | Long protein or large batch | Reduce batch_size |
+| `KeyError: 'LIG'` | Ligand not found in PDB | Check HETATM records |
+| `ValueError: no ligand atoms` | Empty ligand | Verify ligand has atoms in PDB |
+
+---
+
+**Next**: Structure prediction for validation → `protein-qc` for filtering.

package/data/skills/linear-solvers/SKILL.md

@@ -0,0 +1,165 @@
+---
+name: linear-solvers
+description: Select and configure linear solvers for systems Ax=b in dense and sparse problems. Use when choosing direct vs iterative methods, diagnosing convergence issues, estimating conditioning, selecting preconditioners, or debugging stagnation in GMRES/CG/BiCGSTAB.
+allowed-tools: Read, Bash, Write, Grep, Glob
+---
+
+# Linear Solvers
+
+## Goal
+
+Provide a universal workflow to select a solver, assess conditioning, and diagnose convergence for linear systems arising in numerical simulations.
+
+## Requirements
+
+- Python 3.8+
+- NumPy, SciPy (for matrix operations)
+- See individual scripts for dependencies
+
+## Inputs to Gather
+
+| Input | Description | Example |
+|-------|-------------|---------|
+| Matrix size | Dimension of system | `n = 1000000` |
+| Sparsity | Fraction of nonzeros | `0.01%` |
+| Symmetry | Is A = Aᵀ? | `yes` |
+| Definiteness | Is A positive definite? | `yes (SPD)` |
+| Conditioning | Estimated condition number | `10⁶` |
+
+## Decision Guidance
+
+### Solver Selection Flowchart
+
+```
+Is matrix small (n < 5000) and dense?
+├── YES → Use direct solver (LU, Cholesky)
+└── NO → Is matrix symmetric?
+    ├── YES → Is it positive definite?
+    │   ├── YES → Use CG with AMG/IC preconditioner
+    │   └── NO → Use MINRES
+    └── NO → Is it nearly symmetric?
+        ├── YES → Use BiCGSTAB
+        └── NO → Use GMRES with ILU/AMG
+```
+
+### Quick Reference
+
+| Matrix Type | Solver | Preconditioner |
+|-------------|--------|----------------|
+| SPD, sparse | CG | AMG, IC |
+| Symmetric indefinite | MINRES | ILU |
+| Nonsymmetric | GMRES, BiCGSTAB | ILU, AMG |
+| Dense | LU, Cholesky | None |
+| Saddle point | Schur complement, Uzawa | Block preconditioner |
+
+## Script Outputs (JSON Fields)
+
+| Script | Key Outputs |
+|--------|-------------|
+| `scripts/solver_selector.py` | `recommended`, `alternatives`, `notes` |
+| `scripts/convergence_diagnostics.py` | `rate`, `stagnation`, `recommended_action` |
+| `scripts/sparsity_stats.py` | `nnz`, `density`, `bandwidth`, `symmetry` |
+| `scripts/preconditioner_advisor.py` | `suggested`, `notes` |
+| `scripts/scaling_equilibration.py` | `row_scale`, `col_scale`, `notes` |
+| `scripts/residual_norms.py` | `residual_norms`, `relative_norms`, `converged` |
+
+## Workflow
+
+1. **Characterize matrix** - symmetry, definiteness, sparsity
+2. **Analyze sparsity** - Run `scripts/sparsity_stats.py`
+3. **Select solver** - Run `scripts/solver_selector.py`
+4. **Choose preconditioner** - Run `scripts/preconditioner_advisor.py`
+5. **Apply scaling** - If ill-conditioned, use `scripts/scaling_equilibration.py`
+6. **Monitor convergence** - Use `scripts/convergence_diagnostics.py`
+7. **Diagnose issues** - Check residual history with `scripts/residual_norms.py`
+
+## Conversational Workflow Example
+
+**User**: My GMRES solver is stagnating after 50 iterations. The residual drops to 1e-3 then stops improving.
+
+**Agent workflow**:
+1. Diagnose convergence:
+   ```bash
+   python3 scripts/convergence_diagnostics.py --residuals 1,0.1,0.01,0.005,0.003,0.002,0.002,0.002 --json
+   ```
+2. Check for preconditioning advice:
+   ```bash
+   python3 scripts/preconditioner_advisor.py --matrix-type nonsymmetric --sparse --stagnation --json
+   ```
+3. Recommend: Increase restart parameter, try ILU(k) with higher k, or switch to AMG.
+
+## Pre-Solve Checklist
+
+- [ ] Confirm matrix symmetry/definiteness
+- [ ] Decide direct vs iterative based on size and sparsity
+- [ ] Set residual tolerance relative to physics scale
+- [ ] Choose preconditioner appropriate to matrix structure
+- [ ] Apply scaling/equilibration if needed
+- [ ] Track convergence and adjust if stagnation occurs
+
+## CLI Examples
+
+```bash
+# Analyze sparsity pattern
+python3 scripts/sparsity_stats.py --matrix A.npy --json
+
+# Select solver for SPD sparse system
+python3 scripts/solver_selector.py --symmetric --positive-definite --sparse --size 1000000 --json
+
+# Get preconditioner recommendation
+python3 scripts/preconditioner_advisor.py --matrix-type spd --sparse --json
+
+# Diagnose convergence from residual history
+python3 scripts/convergence_diagnostics.py --residuals 1,0.2,0.05,0.01 --json
+
+# Apply scaling
+python3 scripts/scaling_equilibration.py --matrix A.npy --symmetric --json
+
+# Compute residual norms
+python3 scripts/residual_norms.py --residual 1,0.1,0.01 --rhs 1,0,0 --json
+```
+
+## Error Handling
+
+| Error | Cause | Resolution |
+|-------|-------|------------|
+| `Matrix file not found` | Invalid path | Check file exists |
+| `Matrix must be square` | Non-square input | Verify matrix dimensions |
+| `Residuals must be positive` | Invalid residual data | Check input format |
+
+## Interpretation Guidance
+
+### Convergence Rate
+
+| Rate | Meaning | Action |
+|------|---------|--------|
+| < 0.1 | Excellent | Current setup optimal |
+| 0.1 - 0.5 | Good | Acceptable for most problems |
+| 0.5 - 0.9 | Slow | Consider better preconditioner |
+| > 0.9 | Stagnation | Change solver or preconditioner |
+
+### Stagnation Diagnosis
+
+| Pattern | Likely Cause | Fix |
+|---------|--------------|-----|
+| Flat residual | Poor preconditioner | Improve preconditioner |
+| Oscillating | Near-singular or indefinite | Check matrix, try different solver |
+| Very slow decay | Ill-conditioned | Apply scaling, use AMG |
+
+## Limitations
+
+- **Large dense matrices**: Direct solvers may run out of memory
+- **Highly indefinite**: Standard preconditioners may fail
+- **Saddle-point**: Requires specialized block preconditioners
+
+## References
+
+- `references/solver_decision_tree.md` - Selection logic
+- `references/preconditioner_catalog.md` - Preconditioner options
+- `references/convergence_patterns.md` - Diagnosing failures
+- `references/scaling_guidelines.md` - Equilibration guidance
+
+## Version History
+
+- **v1.1.0** (2024-12-24): Enhanced documentation, decision guidance, examples
+- **v1.0.0**: Initial release with 6 solver analysis scripts

package/data/skills/liquid-biopsy-analytics-agent/SKILL.md

@@ -0,0 +1,171 @@
+<!--
+# COPYRIGHT NOTICE
+# This file is part of the "Universal Biomedical Skills" project.
+# Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>
+# All Rights Reserved.
+#
+# This code is proprietary and confidential.
+# Unauthorized copying of this file, via any medium is strictly prohibited.
+#
+# Provenance: Authenticated by MD BABU MIA
+
+-->
+
+---
+name: 'liquid-biopsy-analytics-agent'
+description: 'Comprehensive analysis of liquid biopsy data (ctDNA, CTCs) for cancer detection, MRD monitoring, and response tracking.'
+keywords:
+  - liquid-biopsy
+  - ctdna
+  - mrd
+  - cancer-detection
+  - treatment-response
+measurable_outcome: 'Detects circulating tumor DNA with 0.01% sensitivity and accurately predicts treatment response in longitudinal samples.'
+allowed-tools:
+  - read_file
+  - run_shell_command
+---
+
+
+# Liquid Biopsy Analytics Agent
+
+The **Liquid Biopsy Analytics Agent** provides comprehensive AI-driven analysis of blood-based cancer biomarkers. It integrates circulating tumor DNA (ctDNA), circulating tumor cells (CTCs), exosomes, and cell-free RNA for multi-cancer early detection (MCED), minimal residual disease (MRD) monitoring, and treatment response assessment.
+
+## When to Use This Skill
+
+* For multi-cancer early detection screening from blood samples.
+* To monitor minimal residual disease (MRD) after curative treatment.
+* When tracking tumor evolution and resistance during therapy.
+* For real-time treatment response assessment.
+* To detect cancer recurrence before clinical or imaging evidence.
+
+## Core Capabilities
+
+1. **ctDNA Mutation Analysis**: Variant calling, VAF tracking, and clonal evolution from cell-free DNA.
+
+2. **Methylation-Based Detection**: cfDNA methylation patterns for cancer detection and tissue-of-origin identification.
+
+3. **CTC Enumeration & Analysis**: AI-powered CTC detection, enumeration, and molecular characterization.
+
+4. **Multi-Modal Integration**: Combines ctDNA, CTCs, and protein biomarkers with clinical/imaging data.
+
+5. **MRD Monitoring**: Ultra-sensitive detection of residual disease post-treatment.
+
+6. **Response Prediction**: AI models predicting treatment response from longitudinal liquid biopsy data.
+
+## Analyte Types and Applications
+
+| Analyte | Detection Method | Clinical Use |
+|---------|------------------|--------------|
+| ctDNA mutations | NGS, ddPCR | Therapy selection, resistance |
+| ctDNA methylation | WGBS, targeted | MCED, tissue of origin |
+| ctDNA fragmentation | WGS | Cancer detection |
+| CTCs | CellSearch, microfluidics | Prognosis, monitoring |
+| Exosomes | Immunocapture | Biomarker cargo |
+| cfRNA | RT-qPCR, NGS | Gene expression |
+
+## Workflow
+
+1. **Input**: Liquid biopsy data (ctDNA variants, methylation, CTC counts, protein markers).
+
+2. **Quality Control**: Assess sample quality, input DNA amount, background noise.
+
+3. **Variant Analysis**: Call mutations, calculate VAF, filter artifacts (CHIP).
+
+4. **Multi-analyte Integration**: Combine biomarker signals using ML fusion.
+
+5. **Clinical Interpretation**: Generate actionable insights for treatment decisions.
+
+6. **Longitudinal Tracking**: Model dynamics for response assessment and recurrence detection.
+
+7. **Output**: Cancer detection probability, MRD status, treatment recommendations, clonal evolution.
+
+## Example Usage
+
+**User**: "Analyze longitudinal ctDNA data from this lung cancer patient to assess treatment response and detect resistance."
+
+**Agent Action**:
+```bash
+python3 Skills/Oncology/Liquid_Biopsy_Analytics_Agent/lb_analyzer.py \
+    --ctdna_variants longitudinal_ctdna.vcf \
+    --timepoints week0,week4,week8,week12 \
+    --tumor_markers cea_values.csv \
+    --baseline_tissue baseline_tumor.maf \
+    --analysis response_resistance \
+    --chip_filter true \
+    --output lb_report/
+```
+
+## AI/ML Models
+
+**Multi-Cancer Early Detection (MCED)**:
+- Methylation-based classifiers (sensitivity ~50-80% at 99% specificity)
+- Multi-analyte combination models
+- Tissue-of-origin prediction
+- Integration with imaging and clinical risk
+
+**MRD Detection**:
+- Tumor-informed (personalized panels from tissue)
+- Tumor-agnostic (fixed panels, methylation)
+- Detection limits: 0.01% - 0.001% VAF
+
+**Response Prediction**:
+- Longitudinal VAF dynamics modeling
+- Bayesian evolution frameworks
+- Time-to-progression prediction
+
+## Clonal Hematopoiesis Filtering
+
+Critical challenge in liquid biopsy interpretation:
+
+| Gene | Prevalence | Action |
+|------|------------|--------|
+| DNMT3A | 30-40% of CHIP | Filter if VAF stable, no tumor context |
+| TET2 | 20-30% | Filter if VAF stable |
+| ASXL1 | 10-15% | Filter if VAF stable |
+| TP53 | 5-10% | Context-dependent (tumor vs CHIP) |
+| Matched WBC | Gold standard | Subtract germline/CHIP variants |
+
+## Commercial Platforms (Reference)
+
+| Platform | Technology | Application |
+|----------|------------|-------------|
+| Guardant360 | ctDNA NGS | Therapy selection |
+| FoundationOne Liquid | ctDNA NGS | Comprehensive profiling |
+| Galleri | Methylation | MCED screening |
+| Signatera | Tumor-informed | MRD monitoring |
+| CellSearch | CTC | FDA-cleared enumeration |
+
+## Clinical Decision Points
+
+1. **Treatment Selection**: Actionable mutations (EGFR, ALK, ROS1, BRAF)
+2. **Response Assessment**: ctDNA clearance correlates with outcomes
+3. **Resistance Detection**: Emerging resistance mutations (T790M, C797S)
+4. **Recurrence Monitoring**: Lead time of 3-6 months over imaging
+
+## Prerequisites
+
+* Python 3.10+
+* NGS variant calling pipelines
+* Methylation analysis tools
+* Machine learning frameworks
+
+## Related Skills
+
+* ctDNA_Analysis - For detailed ctDNA workflows
+* Tumor_Clonal_Evolution - For evolutionary analysis
+* MRD_Detection - For residual disease focus
+
+## Limitations and Considerations
+
+- **False positives**: CHIP, benign tumors, inflammation
+- **False negatives**: Low shedding tumors, early stage
+- **Technical variability**: Pre-analytical factors critical
+- **Cost**: Multi-analyte panels expensive
+
+## Author
+
+AI Group - Biomedical AI Platform
+
+
+<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->

package/data/skills/lit-synthesizer/SKILL.md

@@ -0,0 +1,53 @@
+---
+name: lit-synthesizer
+description: Search PubMed and bioRxiv, summarise papers with LLM, build citation graphs, and generate literature review sections.
+version: 0.1.0
+metadata:
+  openclaw:
+    requires:
+      bins:
+        - python3
+      env: []
+      config: []
+    always: false
+    emoji: "🦖"
+    homepage: https://github.com/ClawBio/ClawBio
+    os: [macos, linux]
+    install:
+      - kind: uv
+        package: biopython
+        bins: []
+      - kind: uv
+        package: httpx
+        bins: []
+---
+
+# 🦖 Lit Synthesizer
+
+You are the **Lit Synthesizer**, a specialised agent for biomedical literature search and synthesis.
+
+## Core Capabilities
+
+1. **PubMed Search**: Query NCBI PubMed via Entrez API with MeSH terms
+2. **bioRxiv/medRxiv Search**: Search preprint servers for recent work
+3. **LLM Summarisation**: Summarise abstracts and full texts using the active LLM
+4. **Citation Graph**: Map citation relationships between papers
+5. **Gap Analysis**: Identify understudied areas based on keyword coverage
+6. **Literature Review Drafting**: Generate structured review sections with citations
+
+## Dependencies
+
+- `biopython` (Entrez API access)
+- `httpx` (bioRxiv API)
+- Active LLM for summarisation (uses the agent's own model)
+
+## Example Queries
+
+- "Find the 10 most cited papers on CRISPR in sickle cell disease from 2024-2026"
+- "Summarise recent preprints on ancestry bias in GWAS"
+- "Build a citation graph for genomic equity research"
+- "Draft a literature review paragraph on AlphaFold applications in drug discovery"
+
+## Status
+
+**Planned** -- implementation targeting Week 2-3 (Mar 6-19).