[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EGMRE2F0XF3WSMC4YW9CYW":3,"skill-same:01M7EGMRE2F0XF3WSMC4YW9CYW":81},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":50,"commitSha":51,"skillDirectory":52,"forksCount":14,"repositoryHtmlUrl":53,"repositoryHomepageUrl":53,"ownerHtmlUrl":54,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":55,"install":56,"resources":68},"01M7EGMRE2F0XF3WSMC4YW9CYW","bio-workflows-liquid-biopsy-pipeline",null,"Orchestrates the cell-free DNA \u002F liquid-biopsy pipeline from plasma sequencing to tumor monitoring, forking tumor-naive (screening) vs tumor-informed (MRD), and chaining pre-analytic QC, UMI\u002Fduplex error-suppression (fgbio), fragment QC, ichorCNA tumor fraction (sWGS) or VarDict low-VAF calling (panel), CHIP subtraction against matched WBC, optional fragmentomics\u002Fmethylation, and longitudinal tracking. Use when treating pre-analytics as the irreversible sensitivity ceiling (tube\u002Ftime-to-plasma\u002Fhemolysis), running error-suppression BEFORE calling (single-strand consensus does not remove deamination; only duplex does), reporting a VAF only with input genome-equivalents (TF ~ 2x VAF only for clonal-het-diploid), subtracting CHIP before reporting somatic, or keeping tube\u002Fpanel\u002Fpipeline identical across a longitudinal MRD series. Hands mechanism to the liquid-biopsy component skills; not a re-teach of any single step.","Mr-Nobody-Anonymous\u002FAll-skills","Mr-Nobody-Anonymous","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F184253119?v=4","01M7AEXFJHKTYQW5R0ZCQQTWFV","anonymous","01M7C00M3GQR4QKP9GZA42GFAK",0,"MIT","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-workflows-liquid-biopsy-pipeline\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: bio-workflows-liquid-biopsy-pipeline\ndescription: Orchestrates the cell-free DNA \u002F liquid-biopsy pipeline from plasma sequencing to tumor monitoring, forking tumor-naive (screening) vs tumor-informed (MRD), and chaining pre-analytic QC, UMI\u002Fduplex error-suppression (fgbio), fragment QC, ichorCNA tumor fraction (sWGS) or VarDict low-VAF calling (panel), CHIP subtraction against matched WBC, optional fragmentomics\u002Fmethylation, and longitudinal tracking. Use when treating pre-analytics as the irreversible sensitivity ceiling (tube\u002Ftime-to-plasma\u002Fhemolysis), running error-suppression BEFORE calling (single-strand consensus does not remove deamination; only duplex does), reporting a VAF only with input genome-equivalents (TF ~ 2x VAF only for clonal-het-diploid), subtracting CHIP before reporting somatic, or keeping tube\u002Fpanel\u002Fpipeline identical across a longitudinal MRD series. Hands mechanism to the liquid-biopsy component skills; not a re-teach of any single step.\ntool_type: mixed\nprimary_tool: ichorCNA\ngoal_approach_exempt: true\nworkflow: true\ndepends_on:\n  - liquid-biopsy\u002Fcfdna-preprocessing\n  - liquid-biopsy\u002Fanalytical-validation\n  - liquid-biopsy\u002Fctdna-mutation-detection\n  - liquid-biopsy\u002Ftumor-fraction-estimation\n  - liquid-biopsy\u002Ffragment-analysis\n  - liquid-biopsy\u002Fmethylation-based-detection\n  - liquid-biopsy\u002Flongitudinal-monitoring\nqc_checkpoints:\n  - pre_analytics: \"Tube type\u002Ftime-to-plasma in window; double-spin; low hemolysis; gDNA-contamination fragment check\"\n  - after_consensus: \"Unique-molecule coverage \u002F genome-equivalents recovered (NOT raw depth); duplication plateaued -- read both off the GroupReadsByUmi --family-size-histogram output\"\n  - after_fragment: \"Modal insert ~167bp (150-180); mononucleosome fraction >0.3\"\n  - after_tf: \"TF above the ~3% ichorCNA LoD to trust the value (below = below detection, not low burden)\"\n  - after_mutation: \"VarDict -f 0.005 is a reporting floor, NOT a detection threshold: confirm each call against a per-locus background-error model (PoN or smCounter2) before reporting; CHIP-subtracted against matched WBC\"\nsource: \"https:\u002F\u002Fgithub.com\u002FGPTomics\u002FbioSkills\"\nsource_repository: \"GPTomics\u002FbioSkills\"\nsource_path: \"workflows\u002Fliquid-biopsy-pipeline\u002FSKILL.md\"\nlicense: \"MIT\"\nimported_at: \"2026-09-20\"\n---\n\n## Version Compatibility\n\nReference examples tested with: BWA 0.7.17+, VarDict 1.8+, fgbio 2.1+, ichorCNA 0.6.0+, FinaleToolkit 0.9+ (the `delfi` param was `autosomes` before 0.9, renamed `chrom_sizes`), MethylDackel 0.6+, numpy 1.26+, pandas 2.2+, pysam 0.22+, samtools 1.19+\n\nBefore using code patterns, verify installed versions match. If versions differ:\n- Python: `pip show \u003Cpackage>` then `help(module.function)` to check signatures\n- R: `packageVersion('\u003Cpkg>')` then `?function_name` to verify parameters\n- CLI: `\u003Ctool> --version` then `\u003Ctool> --help` to confirm flags\n\nIf code throws ImportError, AttributeError, or TypeError, introspect the installed\npackage and adapt the example to match the actual API rather than retrying.\n\n# Liquid Biopsy Analysis Pipeline\n\n**\"Analyze my liquid biopsy cfDNA data end-to-end\"** -> Orchestrate UMI-aware preprocessing (fgbio), ctDNA mutation detection (VarDict), tumor fraction estimation (ichorCNA), fragmentomics analysis, and longitudinal monitoring for treatment response.\n\nThis is a workflow skill: it owns the chaining decisions and hand-offs, not the internals of any one step. Every step below cross-references the component skill that teaches its mechanism.\n\n## The governing principle\n\nA liquid-biopsy result is decided at four seams, two of them set before the sequencer ever runs.\n\n1. **Pre-analytics is irreversible and is the sensitivity ceiling — set at the blood draw, not the pipeline.** WBC lysis dumps high-quality germline\u002FCHIP DNA into the denominator; a 0.5% tumor fraction diluted to 0.1% by a processing delay is a false negative with NO bioinformatic recovery. Tube chemistry (Streck ~7d RT vs EDTA \u003C6h), double-spin, and hemolysis are the FIRST gate the whole pipeline is conditional on. Mixing tube types within a longitudinal series is a batch confound.\n2. **The error-suppression \u002F UMI scheme is committed at library prep and cannot be upgraded after sequencing.** Single-strand UMI consensus floors error ~1e-4 to 1e-5 but does NOT remove deamination (C>T) or oxidation (G>T) — those lesions are on the template, so every PCR copy inherits them and the family votes unanimously for the artifact. Only DUPLEX (both-strand concordance) reaches \u003C1e-7. Call on the consensus BAM, never the raw BAM.\n3. **A VAF is undefined without input genome-equivalents, and TF != VAF.** 0.1% on 100 GE is noise; on 30,000 GE it is solid. The LoD is set by input GE and assay design (tumor-informed bespoke integrates 16-50 loci to ppm; tumor-naive fixed panel is error\u002Fsampling-limited ~0.1-0.5%; ichorCNA sWGS floors ~3% TF). TF ~ 2x VAF only for a clonal, heterozygous, diploid locus — copy number breaks the factor of 2.\n4. **Matched buffy-coat\u002FWBC is the CHIP-subtraction commitment.** ~81.6% of cfDNA variants in controls and ~53.2% in cancer patients are CHIP, not tumor (Razavi 2019). Without matched WBC, a tumor-naive cfDNA call is presumptively CHIP-contaminated; a gene-list filter is a weak fallback. Decide at study design.\n\n## Made-once commitments\n\n| Commitment | Consequence inherited downstream |\n|------------|----------------------------------|\n| Pre-analytics (tube\u002Ftime-to-plasma\u002Fdouble-spin) | Irreversible TF dilution; the whole pipeline's sensitivity ceiling; consistent across a longitudinal series |\n| Error-suppression scheme (single-strand vs DUPLEX) | The achievable error floor; single-strand cannot remove deamination\u002Foxidation; cannot upgrade post-hoc |\n| Assay design (tumor-naive vs tumor-informed; panel vs sWGS) + input GE | The LoD and whether it is per-locus or panel-integrated; a VAF is undefined without input GE |\n| Matched WBC available? | Whether CHIP is definitively subtracted (WBC) or only gene-list-filtered (weak) |\n\n## Pipeline Overview\n\n```\nPre-analytical QC -> cfDNA Preprocessing -> Fragment QC\n                          ↓\n        ┌─────────────────┴─────────────────┐\n        ↓                                   ↓\n   sWGS Branch                        Panel Branch\n        ↓                                   ↓\n   ichorCNA                          VarDict\u002FsmCounter2\n   (Tumor Fraction)                  (Mutation Detection)\n        ↓                                   ↓\n        └─────────────────┬─────────────────┘\n                          ↓\n                 Longitudinal Tracking\n```\n\n## Step 0: Pre-Analytical QC\n\n```python\ndef check_preanalytical_quality(sample_metadata):\n    '''\n    Pre-analytical factors critical for cfDNA quality.\n\n    Requirements:\n    - Streck tube: up to 7 days at room temperature\n    - EDTA tube: process within 6 hours\n    - Avoid hemolysis\n    - Store extracted DNA at -80C\n    '''\n    issues = []\n\n    if sample_metadata['tube_type'] == 'EDTA':\n        if sample_metadata['processing_delay_hours'] > 6:\n            issues.append('EDTA tube processed > 6 hours - risk of gDNA contamination')\n\n    if sample_metadata['hemolysis_score'] > 1:\n        issues.append('Hemolysis detected - expect cellular DNA contamination')\n\n    return issues\n```\n\n## Step 1: cfDNA Preprocessing with UMI Consensus\n\n```bash\n# For UMI-tagged libraries (targeted panels)\n# fgbio pipeline\n\n# Extract UMIs. Read-structure is library-specific; see liquid-biopsy\u002Fcfdna-preprocessing.\nfgbio ExtractUmisFromBam \\\n    --input raw.bam \\\n    --output with_umis.bam \\\n    --read-structure 3M2S+T 3M2S+T \\\n    --single-tag RX\n\n# Align. bwa reads FASTQ, not BAM, and stripping to FASTQ drops the RX\u002FUMI tag -- so emit FASTQ\n# carrying RX (samtools fastq -T RX), align, then re-zip the uBAM tags back on with fgbio ZipperBams\n# so GroupReadsByUmi still sees RX.\nsamtools fastq -T RX with_umis.bam | \\\n    bwa mem -C -p -t 8 -Y reference.fa - | \\\n    fgbio ZipperBams --unmapped with_umis.bam --ref reference.fa --output aligned.bam\n\n# Group by UMI. --family-size-histogram is not optional here: rule 3 says a VAF is undefined without\n# input genome-equivalents, and this histogram is where GE and the duplication plateau are read off.\n# Raw depth after consensus collapse is not a sensitivity measure -- unique molecules are.\nfgbio GroupReadsByUmi \\\n    --input aligned.bam \\\n    --output grouped.bam \\\n    --strategy adjacency \\\n    --edits 1 \\\n    --family-size-histogram qc\u002Ffamily_sizes.txt\n\n# Consensus calling: keep the caller permissive (fgbio #1009), apply strictness at the filter\nfgbio CallMolecularConsensusReads \\\n    --input grouped.bam \\\n    --output consensus.bam \\\n    --min-reads 1\n\n# Filter: this is the real quality gate\nfgbio FilterConsensusReads \\\n    --input consensus.bam \\\n    --output final.bam \\\n    --ref reference.fa \\\n    --min-reads 2\n```\n\n## Step 2: Fragment QC Checkpoint\n\n```python\nimport pysam\nimport numpy as np\n\ndef verify_cfdna_quality(bam_path):\n    '''\n    QC Checkpoint: Verify cfDNA fragment profile.\n    Expected: peak at ~167bp (mononucleosome)\n    '''\n    bam = pysam.AlignmentFile(bam_path, 'rb')\n    sizes = []\n\n    for read in bam.fetch():\n        # cfDNA fragments are short; cap at 400 bp (fragments beyond are gDNA\u002Fnoise) so the\n        # modal-size bincount is bounded and not skewed by a few long outliers.\n        if read.is_proper_pair and not read.is_secondary and 0 \u003C read.template_length \u003C= 400:\n            sizes.append(read.template_length)\n\n    bam.close()\n    sizes = np.array(sizes)\n\n    modal_size = np.bincount(sizes).argmax()\n    mono_frac = np.sum((sizes >= 150) & (sizes \u003C= 180)) \u002F len(sizes)\n\n    qc_pass = 150 \u003C= modal_size \u003C= 180 and mono_frac > 0.3\n\n    return {\n        'modal_size': modal_size,\n        'mononucleosome_fraction': mono_frac,\n        'qc_pass': qc_pass,\n        'message': 'Good cfDNA profile' if qc_pass else 'Atypical fragment distribution'\n    }\n```\n\n## Step 3a: Tumor Fraction Estimation (sWGS)\n\nichorCNA is a command-line script (`Rscript scripts\u002FrunIchorCNA.R`), NOT an importable `runIchorCNA()` function, and it is preceded by HMMcopy `readCounter` to bin the BAM. The ~3% tumor-fraction floor is an analytical limit of detection; below it, route to fragmentomics or methylation rather than trusting a low value (see liquid-biopsy\u002Fanalytical-validation and liquid-biopsy\u002Ftumor-fraction-estimation).\n\n```bash\n# For shallow WGS data (0.1-1x coverage); GavinHaLab fork\nreadCounter --window 1000000 --quality 20 \\\n    --chromosome \"chr1,chr2,chr3,chr4,chr5,chr6,chr7,chr8,chr9,chr10,chr11,chr12,chr13,chr14,chr15,chr16,chr17,chr18,chr19,chr20,chr21,chr22,chrX\" \\\n    sample.bam > sample.wig\n\nRscript scripts\u002FrunIchorCNA.R \\\n    --id sample_id --WIG sample.wig \\\n    --gcWig gc_hg38_1000kb.wig --mapWig map_hg38_1000kb.wig \\\n    --centromere GRCh38.GCA_000001405.2_centromere_acen.txt \\\n    --normalPanel HD_ULP_PoN_hg38_1Mb_normAutosomes_median.rds \\   # GavinHaLab\u002FichorCNA extdata name; PoN build MUST match the gc\u002Fmap\u002Fcentromere build (the non-hg38 1Mb PoN is hg19)\n    --normal \"c(0.5,0.6,0.7,0.8,0.9)\" --ploidy \"c(2,3)\" --maxCN 7 \\\n    --estimateNormal TRUE --estimatePloidy TRUE --estimateScPrevalence TRUE \\\n    --outDir ichor_results\u002F\n# Tumor fraction = 1 - n in sample_id.params.txt\n```\n\n## Step 3b: Mutation Detection (Targeted Panel)\n\n```bash\n# For deep targeted sequencing\n# Use UMI-consensus BAM from Step 1\n\nvardict-java \\\n    -G reference.fa \\\n    -f 0.005 \\\n    -N sample_id \\\n    -b consensus.bam \\\n    -c 1 -S 2 -E 3 -g 4 \\\n    panel.bed | \\\nteststrandbias.R | \\\nvar2vcf_valid.pl \\\n    -N sample_id \\\n    -E \\\n    -f 0.005 \\\n    > sample.vcf\n```\n\n## Step 4: CHIP Filtering\n\nClonal hematopoiesis (CHIP) is the dominant false-positive source in plasma: ~81.6% of cfDNA variants in controls and ~53.2% in cancer patients trace to white blood cells (Razavi 2019 Nat Med 25:1928). A gene-list filter is a weak fallback; the definitive control is sequencing matched buffy-coat\u002FWBC DNA and subtracting any variant present there. See liquid-biopsy\u002Fctdna-mutation-detection.\n\n```python\nCHIP_GENES = ['DNMT3A', 'TET2', 'ASXL1', 'PPM1D', 'JAK2', 'SF3B1', 'SRSF2', 'TP53']\n\ndef filter_chip(variants_df, wbc_variants=None, chip_genes=CHIP_GENES):\n    '''Subtract WBC-matched variants when available; else fall back to a CHIP gene list.'''\n    if wbc_variants is not None:\n        # REF must be in the key. Left-aligned indels share (chrom, pos, alt): a 1bp deletion\n        # (REF=AT, ALT=A) and a 2bp deletion (REF=ATT, ALT=A) collide, so a (chrom, pos, alt) key\n        # would silently subtract a real somatic indel as CHIP.\n        key = ['chrom', 'pos', 'ref', 'alt']\n        wbc_keys = set(map(tuple, wbc_variants[key].itertuples(index=False, name=None)))\n        in_wbc = variants_df[key].apply(lambda r: tuple(r) in wbc_keys, axis=1)\n        return variants_df[~in_wbc], variants_df[in_wbc]\n\n    chip = variants_df[variants_df['gene'].isin(chip_genes)]\n    somatic = variants_df[~variants_df['gene'].isin(chip_genes)]\n    return somatic, chip\n```\n\n## Step 5: Fragmentomics Analysis (Optional)\n\nFinaleToolkit (MIT license, not DELFI software) exposes real hyphenated CLI subcommands and an underscored `finaletoolkit.frag` Python API; `delfi` GC-corrects the short\u002Flong ratio (raw ratios are dominated by GC and sequencing batch). DELFI is a methodology and a company, not a `pip install`-able tool.\n\n```bash\n# GC-corrected genome-wide DELFI profile and end-motif diversity.\n# delfi positionals: input chrom_sizes reference bins_file; GC correction is on by default.\n# --no-remove-nocov keeps all bins on non-hg19 references (the default removes two hardcoded hg19 no-coverage regions).\nfinaletoolkit delfi consensus.bam hg38.chrom.sizes hg38.2bit bins_100kb.bed -g gaps.bed --no-remove-nocov -o sample.delfi.bed\nfinaletoolkit end-motifs consensus.bam hg38.2bit -o sample.end_motifs.tsv\nfinaletoolkit mds sample.end_motifs.tsv\n```\n\n```python\nfrom finaletoolkit.frag import delfi  # see liquid-biopsy\u002Ffragment-analysis\n\ndef run_fragmentomics(bam_path, chrom_sizes, reference, bins_bed, gap_bed):\n    '''GC-corrected DELFI short\u002Flong profile (MDS comes from end_motifs().motif_diversity_score()).\n    Python positional order is (input, chrom_sizes, bins_file, reference_file) - note this differs\n    from the CLI order (input, chrom_sizes, reference, bins), so pass by keyword to be safe.'''\n    return delfi(bam_path, chrom_sizes=chrom_sizes, bins_file=bins_bed,\n                 reference_file=reference, gap_file=gap_bed)\n```\n\n## Step 6: Longitudinal Tracking\n\n```python\nimport pandas as pd\nimport numpy as np\n\ndef track_longitudinal(samples_df):\n    '''\n    Track ctDNA over treatment.\n\n    samples_df columns: [sample_id, timepoint, tumor_fraction, mutations...]\n    '''\n    samples_df = samples_df.sort_values('timepoint')\n\n    baseline = samples_df.iloc[0]['tumor_fraction']\n    samples_df['log2_fc'] = np.log2(samples_df['tumor_fraction'] \u002F baseline)\n\n    nadir = samples_df['tumor_fraction'].min()\n\n    response = 'unknown'\n    if nadir \u003C 0.001:\n        response = 'Complete molecular response'\n    elif nadir \u003C baseline * 0.01:\n        response = 'Major molecular response (>2 log)'\n    elif nadir \u003C baseline * 0.5:\n        response = 'Partial molecular response'\n\n    return samples_df, response\n```\n\n## Complete Pipeline Script\n\n```python\ndef run_liquid_biopsy_pipeline(sample_config):\n    '''\n    Complete liquid biopsy analysis pipeline.\n\n    sample_config: dict with keys:\n        - bam_file: Input BAM\n        - data_type: 'swgs' or 'panel'\n        - reference: Reference FASTA\n        - bed_file: Panel BED (for panel data)\n        - output_dir: Output directory\n    '''\n    results = {}\n\n    # Step 1: Preprocess (if UMI data)\n    if sample_config.get('has_umis'):\n        preprocessed_bam = preprocess_with_fgbio(sample_config['bam_file'])\n    else:\n        preprocessed_bam = sample_config['bam_file']\n\n    # Step 2: Fragment QC\n    frag_qc = verify_cfdna_quality(preprocessed_bam)\n    if not frag_qc['qc_pass']:\n        print(f\"WARNING: {frag_qc['message']}\")\n    results['fragment_qc'] = frag_qc\n\n    # Step 3: Analysis based on data type\n    if sample_config['data_type'] == 'swgs':\n        # Tumor fraction estimation\n        results['tumor_fraction'] = run_ichorcna(preprocessed_bam)\n    elif sample_config['data_type'] == 'panel':\n        # Mutation detection. CHIP subtraction is not optional (rule 4): without matched WBC the\n        # calls are presumptively CHIP-contaminated and the gene list is only a weak fallback.\n        variants = call_variants(preprocessed_bam, sample_config['bed_file'])\n        wbc = call_variants(sample_config['wbc_bam'], sample_config['bed_file']) if sample_config.get('wbc_bam') else None\n        somatic, chip = filter_chip(variants, wbc_variants=wbc)\n        results['variants'] = somatic\n        results['chip_variants'] = chip\n\n    # Step 4: Optional fragmentomics\n    if sample_config.get('run_fragmentomics'):\n        results['fragmentomics'] = run_fragmentomics(preprocessed_bam)\n\n    return results\n```\n\n## Choosing the branch (tumor-naive vs tumor-informed)\n\nPipeline-level selection only; mechanism lives in the component skills.\n\n| Situation | Branch | Hand off to |\n|-----------|--------|-------------|\n| Screening \u002F no known tumor \u002F MCED | tumor-NAIVE (fixed panel, sWGS, or methylation) | liquid-biopsy\u002Fmethylation-based-detection |\n| MRD\u002Frecurrence of a KNOWN tumor | tumor-INFORMED bespoke (16-50 clonal variants from tissue WES) | liquid-biopsy\u002Flongitudinal-monitoring |\n| Tumor fraction from sWGS (0.1-1x) | ichorCNA (copy-number-based, floor ~3% TF) | liquid-biopsy\u002Ftumor-fraction-estimation |\n| Low-VAF mutations from a deep panel | VarDict \u002F smCounter2 on the UMI-consensus BAM | liquid-biopsy\u002Fctdna-mutation-detection |\n| Below mutation LoD, still need signal | fragmentomics (DELFI\u002Fend-motifs) or methylation | liquid-biopsy\u002Ffragment-analysis |\n| Stating\u002Ftrusting a sensitivity claim | LoB\u002FLoD\u002FLoD95\u002FLoQ, per-locus vs panel-integrated | liquid-biopsy\u002Fanalytical-validation |\n\nTumor-informed reaches ppm by integrating across 16-50 loci (beats the single-locus Poisson floor) but is BLIND by design to new\u002Fresistance variants; tumor-naive sees any panel variant but is CHIP-dominated and error-limited.\n\n## Common Errors\n\n| Symptom | Cause | Fix |\n|---------|-------|-----|\n| False negative on a real low-TF sample | Bad blood draw irreversibly diluted TF | Pre-analytic gate; cannot be rescued downstream |\n| Residual C>T \u002F G>T false positives | Single-strand consensus assumed to remove damage | Duplex, or UDG for deamination; flag the substitution spectrum |\n| Money spent, no sensitivity gain | Sequenced past molecular saturation | Report unique-molecule coverage; invest in plasma volume \u002F conversion efficiency |\n| Blood clones reported as tumor | CHIP not subtracted (esp. TP53\u002FPPM1D, both CHIP and driver) | Matched buffy-coat\u002FWBC sequencing (gene list is a weak fallback) |\n| Apparent burden halved or doubled | VAF reported as TF (or vice versa) | State VAF vs TF; TF ~ 2x VAF only for clonal-het-diploid; prefer CNA-based TF when CN is non-neutral |\n| A \"low\" TF trusted below the assay floor | ~3% ichorCNA LoD ignored | Below the floor = below detection; route to fragmentomics\u002Fmethylation |\n| Longitudinal \"response\" that is a batch shift | Tube\u002Fpanel\u002Fpipeline changed between draws | Same tube, same panel, same pipeline across timepoints |\n\n## References\n\n- Razavi P, Li BT, Brown DN, et al (2019) High-intensity sequencing reveals the sources of plasma circulating cell-free DNA variants. *Nature Medicine* 25:1928-1937. DOI 10.1038\u002Fs41591-019-0652-7. (CHIP dominance of cfDNA variants.)\n- Adalsteinsson VA, Ha G, Freeman SS, et al (2017) Scalable whole-exome sequencing of cell-free DNA reveals high concordance with metastatic tumors. *Nature Communications* 8:1324. DOI 10.1038\u002Fs41467-017-00965-y. (ichorCNA; ~3% TF LoD.)\n- Schmitt MW, Kennedy SR, Salk JJ, et al (2012) Detection of ultra-rare mutations by next-generation sequencing. *PNAS* 109:14508-14513. DOI 10.1073\u002Fpnas.1208715109. (duplex error floor.)\n- Reinert T, Henriksen TV, Christensen E, et al (2019) Analysis of plasma cell-free DNA by ultradeep sequencing in patients with stages I to III colorectal cancer. *JAMA Oncology* 5:1124-1131. DOI 10.1001\u002Fjamaoncol.2019.0528. (tumor-informed bespoke MRD.)\n\n## Related Skills\n\n- liquid-biopsy\u002Fcfdna-preprocessing - UMI\u002Fduplex consensus error suppression\n- liquid-biopsy\u002Fanalytical-validation - molecule-counting limits of detection and honest LoD reporting\n- liquid-biopsy\u002Fctdna-mutation-detection - low-VAF calling and CHIP subtraction\n- liquid-biopsy\u002Ftumor-fraction-estimation - ichorCNA tumor fraction from sWGS\n- liquid-biopsy\u002Ffragment-analysis - fragmentomics features\n- liquid-biopsy\u002Fmethylation-based-detection - methylation detection and tissue-of-origin\n- liquid-biopsy\u002Flongitudinal-monitoring - serial MRD tracking\n","## Version Compatibility\n\nReference examples tested with: BWA 0.7.17+, VarDict 1.8+, fgbio 2.1+, ichorCNA 0.6.0+, FinaleToolkit 0.9+ (the `delfi` param was `autosomes` before 0.9, renamed `chrom_sizes`), MethylDackel 0.6+, numpy 1.26+, pandas 2.2+, pysam 0.22+, samtools 1.19+\n\nBefore using code patterns, verify installed versions match. If versions differ:\n- Python: `pip show \u003Cpackage>` then `help(module.function)` to check signatures\n- R: `packageVersion('\u003Cpkg>')` then `?function_name` to verify parameters\n- CLI: `\u003Ctool> --version` then `\u003Ctool> --help` to confirm flags\n\nIf code throws ImportError, AttributeError, or TypeError, introspect the installed\npackage and adapt the example to match the actual API rather than retrying.\n\n# Liquid Biopsy Analysis Pipeline\n\n**\"Analyze my liquid biopsy cfDNA data end-to-end\"** -> Orchestrate UMI-aware preprocessing (fgbio), ctDNA mutation detection (VarDict), tumor fraction estimation (ichorCNA), fragmentomics analysis, and longitudinal monitoring for treatment response.\n\nThis is a workflow skill: it owns the chaining decisions and hand-offs, not the internals of any one step. Every step below cross-references the component skill that teaches its mechanism.\n\n## The governing principle\n\nA liquid-biopsy result is decided at four seams, two of them set before the sequencer ever runs.\n\n1. **Pre-analytics is irreversible and is the sensitivity ceiling — set at the blood draw, not the pipeline.** WBC lysis dumps high-quality germline\u002FCHIP DNA into the denominator; a 0.5% tumor fraction diluted to 0.1% by a processing delay is a false negative with NO bioinformatic recovery. Tube chemistry (Streck ~7d RT vs EDTA \u003C6h), double-spin, and hemolysis are the FIRST gate the whole pipeline is conditional on. Mixing tube types within a longitudinal series is a batch confound.\n2. **The error-suppression \u002F UMI scheme is committed at library prep and cannot be upgraded after sequencing.** Single-strand UMI consensus floors error ~1e-4 to 1e-5 but does NOT remove deamination (C>T) or oxidation (G>T) — those lesions are on the template, so every PCR copy inherits them and the family votes unanimously for the artifact. Only DUPLEX (both-strand concordance) reaches \u003C1e-7. Call on the consensus BAM, never the raw BAM.\n3. **A VAF is undefined without input genome-equivalents, and TF != VAF.** 0.1% on 100 GE is noise; on 30,000 GE it is solid. The LoD is set by input GE and assay design (tumor-informed bespoke integrates 16-50 loci to ppm; tumor-naive fixed panel is error\u002Fsampling-limited ~0.1-0.5%; ichorCNA sWGS floors ~3% TF). TF ~ 2x VAF only for a clonal, heterozygous, diploid locus — copy number breaks the factor of 2.\n4. **Matched buffy-coat\u002FWBC is the CHIP-subtraction commitment.** ~81.6% of cfDNA variants in controls and ~53.2% in cancer patients are CHIP, not tumor (Razavi 2019). Without matched WBC, a tumor-naive cfDNA call is presumptively CHIP-contaminated; a gene-list filter is a weak fallback. Decide at study design.\n\n## Made-once commitments\n\n| Commitment | Consequence inherited downstream |\n|------------|----------------------------------|\n| Pre-analytics (tube\u002Ftime-to-plasma\u002Fdouble-spin) | Irreversible TF dilution; the whole pipeline's sensitivity ceiling; consistent across a longitudinal series |\n| Error-suppression scheme (single-strand vs DUPLEX) | The achievable error floor; single-strand cannot remove deamination\u002Foxidation; cannot upgrade post-hoc |\n| Assay design (tumor-naive vs tumor-informed; panel vs sWGS) + input GE | The LoD and whether it is per-locus or panel-integrated; a VAF is undefined without input GE |\n| Matched WBC available? | Whether CHIP is definitively subtracted (WBC) or only gene-list-filtered (weak) |\n\n## Pipeline Overview\n\n```\nPre-analytical QC -> cfDNA Preprocessing -> Fragment QC\n                          ↓\n        ┌─────────────────┴─────────────────┐\n        ↓                                   ↓\n   sWGS Branch                        Panel Branch\n        ↓                                   ↓\n   ichorCNA                          VarDict\u002FsmCounter2\n   (Tumor Fraction)                  (Mutation Detection)\n        ↓                                   ↓\n        └─────────────────┬─────────────────┘\n                          ↓\n                 Longitudinal Tracking\n```\n\n## Step 0: Pre-Analytical QC\n\n```python\ndef check_preanalytical_quality(sample_metadata):\n    '''\n    Pre-analytical factors critical for cfDNA quality.\n\n    Requirements:\n    - Streck tube: up to 7 days at room temperature\n    - EDTA tube: process within 6 hours\n    - Avoid hemolysis\n    - Store extracted DNA at -80C\n    '''\n    issues = []\n\n    if sample_metadata['tube_type'] == 'EDTA':\n        if sample_metadata['processing_delay_hours'] > 6:\n            issues.append('EDTA tube processed > 6 hours - risk of gDNA contamination')\n\n    if sample_metadata['hemolysis_score'] > 1:\n        issues.append('Hemolysis detected - expect cellular DNA contamination')\n\n    return issues\n```\n\n## Step 1: cfDNA Preprocessing with UMI Consensus\n\n```bash\n# For UMI-tagged libraries (targeted panels)\n# fgbio pipeline\n\n# Extract UMIs. Read-structure is library-specific; see liquid-biopsy\u002Fcfdna-preprocessing.\nfgbio ExtractUmisFromBam \\\n    --input raw.bam \\\n    --output with_umis.bam \\\n    --read-structure 3M2S+T 3M2S+T \\\n    --single-tag RX\n\n# Align. bwa reads FASTQ, not BAM, and stripping to FASTQ drops the RX\u002FUMI tag -- so emit FASTQ\n# carrying RX (samtools fastq -T RX), align, then re-zip the uBAM tags back on with fgbio ZipperBams\n# so GroupReadsByUmi still sees RX.\nsamtools fastq -T RX with_umis.bam | \\\n    bwa mem -C -p -t 8 -Y reference.fa - | \\\n    fgbio ZipperBams --unmapped with_umis.bam --ref reference.fa --output aligned.bam\n\n# Group by UMI. --family-size-histogram is not optional here: rule 3 says a VAF is undefined without\n# input genome-equivalents, and this histogram is where GE and the duplication plateau are read off.\n# Raw depth after consensus collapse is not a sensitivity measure -- unique molecules are.\nfgbio GroupReadsByUmi \\\n    --input aligned.bam \\\n    --output grouped.bam \\\n    --strategy adjacency \\\n    --edits 1 \\\n    --family-size-histogram qc\u002Ffamily_sizes.txt\n\n# Consensus calling: keep the caller permissive (fgbio #1009), apply strictness at the filter\nfgbio CallMolecularConsensusReads \\\n    --input grouped.bam \\\n    --output consensus.bam \\\n    --min-reads 1\n\n# Filter: this is the real quality gate\nfgbio FilterConsensusReads \\\n    --input consensus.bam \\\n    --output final.bam \\\n    --ref reference.fa \\\n    --min-reads 2\n```\n\n## Step 2: Fragment QC Checkpoint\n\n```python\nimport pysam\nimport numpy as np\n\ndef verify_cfdna_quality(bam_path):\n    '''\n    QC Checkpoint: Verify cfDNA fragment profile.\n    Expected: peak at ~167bp (mononucleosome)\n    '''\n    bam = pysam.AlignmentFile(bam_path, 'rb')\n    sizes = []\n\n    for read in bam.fetch():\n        # cfDNA fragments are short; cap at 400 bp (fragments beyond are gDNA\u002Fnoise) so the\n        # modal-size bincount is bounded and not skewed by a few long outliers.\n        if read.is_proper_pair and not read.is_secondary and 0 \u003C read.template_length \u003C= 400:\n            sizes.append(read.template_length)\n\n    bam.close()\n    sizes = np.array(sizes)\n\n    modal_size = np.bincount(sizes).argmax()\n    mono_frac = np.sum((sizes >= 150) & (sizes \u003C= 180)) \u002F len(sizes)\n\n    qc_pass = 150 \u003C= modal_size \u003C= 180 and mono_frac > 0.3\n\n    return {\n        'modal_size': modal_size,\n        'mononucleosome_fraction': mono_frac,\n        'qc_pass': qc_pass,\n        'message': 'Good cfDNA profile' if qc_pass else 'Atypical fragment distribution'\n    }\n```\n\n## Step 3a: Tumor Fraction Estimation (sWGS)\n\nichorCNA is a command-line script (`Rscript scripts\u002FrunIchorCNA.R`), NOT an importable `runIchorCNA()` function, and it is preceded by HMMcopy `readCounter` to bin the BAM. The ~3% tumor-fraction floor is an analytical limit of detection; below it, route to fragmentomics or methylation rather than trusting a low value (see liquid-biopsy\u002Fanalytical-validation and liquid-biopsy\u002Ftumor-fraction-estimation).\n\n```bash\n# For shallow WGS data (0.1-1x coverage); GavinHaLab fork\nreadCounter --window 1000000 --quality 20 \\\n    --chromosome \"chr1,chr2,chr3,chr4,chr5,chr6,chr7,chr8,chr9,chr10,chr11,chr12,chr13,chr14,chr15,chr16,chr17,chr18,chr19,chr20,chr21,chr22,chrX\" \\\n    sample.bam > sample.wig\n\nRscript scripts\u002FrunIchorCNA.R \\\n    --id sample_id --WIG sample.wig \\\n    --gcWig gc_hg38_1000kb.wig --mapWig map_hg38_1000kb.wig \\\n    --centromere GRCh38.GCA_000001405.2_centromere_acen.txt \\\n    --normalPanel HD_ULP_PoN_hg38_1Mb_normAutosomes_median.rds \\   # GavinHaLab\u002FichorCNA extdata name; PoN build MUST match the gc\u002Fmap\u002Fcentromere build (the non-hg38 1Mb PoN is hg19)\n    --normal \"c(0.5,0.6,0.7,0.8,0.9)\" --ploidy \"c(2,3)\" --maxCN 7 \\\n    --estimateNormal TRUE --estimatePloidy TRUE --estimateScPrevalence TRUE \\\n    --outDir ichor_results\u002F\n# Tumor fraction = 1 - n in sample_id.params.txt\n```\n\n## Step 3b: Mutation Detection (Targeted Panel)\n\n```bash\n# For deep targeted sequencing\n# Use UMI-consensus BAM from Step 1\n\nvardict-java \\\n    -G reference.fa \\\n    -f 0.005 \\\n    -N sample_id \\\n    -b consensus.bam \\\n    -c 1 -S 2 -E 3 -g 4 \\\n    panel.bed | \\\nteststrandbias.R | \\\nvar2vcf_valid.pl \\\n    -N sample_id \\\n    -E \\\n    -f 0.005 \\\n    > sample.vcf\n```\n\n## Step 4: CHIP Filtering\n\nClonal hematopoiesis (CHIP) is the dominant false-positive source in plasma: ~81.6% of cfDNA variants in controls and ~53.2% in cancer patients trace to white blood cells (Razavi 2019 Nat Med 25:1928). A gene-list filter is a weak fallback; the definitive control is sequencing matched buffy-coat\u002FWBC DNA and subtracting any variant present there. See liquid-biopsy\u002Fctdna-mutation-detection.\n\n```python\nCHIP_GENES = ['DNMT3A', 'TET2', 'ASXL1', 'PPM1D', 'JAK2', 'SF3B1', 'SRSF2', 'TP53']\n\ndef filter_chip(variants_df, wbc_variants=None, chip_genes=CHIP_GENES):\n    '''Subtract WBC-matched variants when available; else fall back to a CHIP gene list.'''\n    if wbc_variants is not None:\n        # REF must be in the key. Left-aligned indels share (chrom, pos, alt): a 1bp deletion\n        # (REF=AT, ALT=A) and a 2bp deletion (REF=ATT, ALT=A) collide, so a (chrom, pos, alt) key\n        # would silently subtract a real somatic indel as CHIP.\n        key = ['chrom', 'pos', 'ref', 'alt']\n        wbc_keys = set(map(tuple, wbc_variants[key].itertuples(index=False, name=None)))\n        in_wbc = variants_df[key].apply(lambda r: tuple(r) in wbc_keys, axis=1)\n        return variants_df[~in_wbc], variants_df[in_wbc]\n\n    chip = variants_df[variants_df['gene'].isin(chip_genes)]\n    somatic = variants_df[~variants_df['gene'].isin(chip_genes)]\n    return somatic, chip\n```\n\n## Step 5: Fragmentomics Analysis (Optional)\n\nFinaleToolkit (MIT license, not DELFI software) exposes real hyphenated CLI subcommands and an underscored `finaletoolkit.frag` Python API; `delfi` GC-corrects the short\u002Flong ratio (raw ratios are dominated by GC and sequencing batch). DELFI is a methodology and a company, not a `pip install`-able tool.\n\n```bash\n# GC-corrected genome-wide DELFI profile and end-motif diversity.\n# delfi positionals: input chrom_sizes reference bins_file; GC correction is on by default.\n# --no-remove-nocov keeps all bins on non-hg19 references (the default removes two hardcoded hg19 no-coverage regions).\nfinaletoolkit delfi consensus.bam hg38.chrom.sizes hg38.2bit bins_100kb.bed -g gaps.bed --no-remove-nocov -o sample.delfi.bed\nfinaletoolkit end-motifs consensus.bam hg38.2bit -o sample.end_motifs.tsv\nfinaletoolkit mds sample.end_motifs.tsv\n```\n\n```python\nfrom finaletoolkit.frag import delfi  # see liquid-biopsy\u002Ffragment-analysis\n\ndef run_fragmentomics(bam_path, chrom_sizes, reference, bins_bed, gap_bed):\n    '''GC-corrected DELFI short\u002Flong profile (MDS comes from end_motifs().motif_diversity_score()).\n    Python positional order is (input, chrom_sizes, bins_file, reference_file) - note this differs\n    from the CLI order (input, chrom_sizes, reference, bins), so pass by keyword to be safe.'''\n    return delfi(bam_path, chrom_sizes=chrom_sizes, bins_file=bins_bed,\n                 reference_file=reference, gap_file=gap_bed)\n```\n\n## Step 6: Longitudinal Tracking\n\n```python\nimport pandas as pd\nimport numpy as np\n\ndef track_longitudinal(samples_df):\n    '''\n    Track ctDNA over treatment.\n\n    samples_df columns: [sample_id, timepoint, tumor_fraction, mutations...]\n    '''\n    samples_df = samples_df.sort_values('timepoint')\n\n    baseline = samples_df.iloc[0]['tumor_fraction']\n    samples_df['log2_fc'] = np.log2(samples_df['tumor_fraction'] \u002F baseline)\n\n    nadir = samples_df['tumor_fraction'].min()\n\n    response = 'unknown'\n    if nadir \u003C 0.001:\n        response = 'Complete molecular response'\n    elif nadir \u003C baseline * 0.01:\n        response = 'Major molecular response (>2 log)'\n    elif nadir \u003C baseline * 0.5:\n        response = 'Partial molecular response'\n\n    return samples_df, response\n```\n\n## Complete Pipeline Script\n\n```python\ndef run_liquid_biopsy_pipeline(sample_config):\n    '''\n    Complete liquid biopsy analysis pipeline.\n\n    sample_config: dict with keys:\n        - bam_file: Input BAM\n        - data_type: 'swgs' or 'panel'\n        - reference: Reference FASTA\n        - bed_file: Panel BED (for panel data)\n        - output_dir: Output directory\n    '''\n    results = {}\n\n    # Step 1: Preprocess (if UMI data)\n    if sample_config.get('has_umis'):\n        preprocessed_bam = preprocess_with_fgbio(sample_config['bam_file'])\n    else:\n        preprocessed_bam = sample_config['bam_file']\n\n    # Step 2: Fragment QC\n    frag_qc = verify_cfdna_quality(preprocessed_bam)\n    if not frag_qc['qc_pass']:\n        print(f\"WARNING: {frag_qc['message']}\")\n    results['fragment_qc'] = frag_qc\n\n    # Step 3: Analysis based on data type\n    if sample_config['data_type'] == 'swgs':\n        # Tumor fraction estimation\n        results['tumor_fraction'] = run_ichorcna(preprocessed_bam)\n    elif sample_config['data_type'] == 'panel':\n        # Mutation detection. CHIP subtraction is not optional (rule 4): without matched WBC the\n        # calls are presumptively CHIP-contaminated and the gene list is only a weak fallback.\n        variants = call_variants(preprocessed_bam, sample_config['bed_file'])\n        wbc = call_variants(sample_config['wbc_bam'], sample_config['bed_file']) if sample_config.get('wbc_bam') else None\n        somatic, chip = filter_chip(variants, wbc_variants=wbc)\n        results['variants'] = somatic\n        results['chip_variants'] = chip\n\n    # Step 4: Optional fragmentomics\n    if sample_config.get('run_fragmentomics'):\n        results['fragmentomics'] = run_fragmentomics(preprocessed_bam)\n\n    return results\n```\n\n## Choosing the branch (tumor-naive vs tumor-informed)\n\nPipeline-level selection only; mechanism lives in the component skills.\n\n| Situation | Branch | Hand off to |\n|-----------|--------|-------------|\n| Screening \u002F no known tumor \u002F MCED | tumor-NAIVE (fixed panel, sWGS, or methylation) | liquid-biopsy\u002Fmethylation-based-detection |\n| MRD\u002Frecurrence of a KNOWN tumor | tumor-INFORMED bespoke (16-50 clonal variants from tissue WES) | liquid-biopsy\u002Flongitudinal-monitoring |\n| Tumor fraction from sWGS (0.1-1x) | ichorCNA (copy-number-based, floor ~3% TF) | liquid-biopsy\u002Ftumor-fraction-estimation |\n| Low-VAF mutations from a deep panel | VarDict \u002F smCounter2 on the UMI-consensus BAM | liquid-biopsy\u002Fctdna-mutation-detection |\n| Below mutation LoD, still need signal | fragmentomics (DELFI\u002Fend-motifs) or methylation | liquid-biopsy\u002Ffragment-analysis |\n| Stating\u002Ftrusting a sensitivity claim | LoB\u002FLoD\u002FLoD95\u002FLoQ, per-locus vs panel-integrated | liquid-biopsy\u002Fanalytical-validation |\n\nTumor-informed reaches ppm by integrating across 16-50 loci (beats the single-locus Poisson floor) but is BLIND by design to new\u002Fresistance variants; tumor-naive sees any panel variant but is CHIP-dominated and error-limited.\n\n## Common Errors\n\n| Symptom | Cause | Fix |\n|---------|-------|-----|\n| False negative on a real low-TF sample | Bad blood draw irreversibly diluted TF | Pre-analytic gate; cannot be rescued downstream |\n| Residual C>T \u002F G>T false positives | Single-strand consensus assumed to remove damage | Duplex, or UDG for deamination; flag the substitution spectrum |\n| Money spent, no sensitivity gain | Sequenced past molecular saturation | Report unique-molecule coverage; invest in plasma volume \u002F conversion efficiency |\n| Blood clones reported as tumor | CHIP not subtracted (esp. TP53\u002FPPM1D, both CHIP and driver) | Matched buffy-coat\u002FWBC sequencing (gene list is a weak fallback) |\n| Apparent burden halved or doubled | VAF reported as TF (or vice versa) | State VAF vs TF; TF ~ 2x VAF only for clonal-het-diploid; prefer CNA-based TF when CN is non-neutral |\n| A \"low\" TF trusted below the assay floor | ~3% ichorCNA LoD ignored | Below the floor = below detection; route to fragmentomics\u002Fmethylation |\n| Longitudinal \"response\" that is a batch shift | Tube\u002Fpanel\u002Fpipeline changed between draws | Same tube, same panel, same pipeline across timepoints |\n\n## References\n\n- Razavi P, Li BT, Brown DN, et al (2019) High-intensity sequencing reveals the sources of plasma circulating cell-free DNA variants. *Nature Medicine* 25:1928-1937. DOI 10.1038\u002Fs41591-019-0652-7. (CHIP dominance of cfDNA variants.)\n- Adalsteinsson VA, Ha G, Freeman SS, et al (2017) Scalable whole-exome sequencing of cell-free DNA reveals high concordance with metastatic tumors. *Nature Communications* 8:1324. DOI 10.1038\u002Fs41467-017-00965-y. (ichorCNA; ~3% TF LoD.)\n- Schmitt MW, Kennedy SR, Salk JJ, et al (2012) Detection of ultra-rare mutations by next-generation sequencing. *PNAS* 109:14508-14513. DOI 10.1073\u002Fpnas.1208715109. (duplex error floor.)\n- Reinert T, Henriksen TV, Christensen E, et al (2019) Analysis of plasma cell-free DNA by ultradeep sequencing in patients with stages I to III colorectal cancer. *JAMA Oncology* 5:1124-1131. DOI 10.1001\u002Fjamaoncol.2019.0528. (tumor-informed bespoke MRD.)\n\n## Related Skills\n\n- liquid-biopsy\u002Fcfdna-preprocessing - UMI\u002Fduplex consensus error suppression\n- liquid-biopsy\u002Fanalytical-validation - molecule-counting limits of detection and honest LoD reporting\n- liquid-biopsy\u002Fctdna-mutation-detection - low-VAF calling and CHIP subtraction\n- liquid-biopsy\u002Ftumor-fraction-estimation - ichorCNA tumor fraction from sWGS\n- liquid-biopsy\u002Ffragment-analysis - fragmentomics features\n- liquid-biopsy\u002Fmethylation-based-detection - methylation detection and tissue-of-origin\n- liquid-biopsy\u002Flongitudinal-monitoring - serial MRD tracking\n",{"name":5,"source":24,"license":15,"workflow":25,"tool_type":26,"depends_on":27,"description":7,"imported_at":35,"source_path":36,"primary_tool":37,"qc_checkpoints":38,"source_repository":49,"goal_approach_exempt":25},"https:\u002F\u002Fgithub.com\u002FGPTomics\u002FbioSkills",true,"mixed",[28,29,30,31,32,33,34],"liquid-biopsy\u002Fcfdna-preprocessing","liquid-biopsy\u002Fanalytical-validation","liquid-biopsy\u002Fctdna-mutation-detection","liquid-biopsy\u002Ftumor-fraction-estimation","liquid-biopsy\u002Ffragment-analysis","liquid-biopsy\u002Fmethylation-based-detection","liquid-biopsy\u002Flongitudinal-monitoring","2026-09-20","workflows\u002Fliquid-biopsy-pipeline\u002FSKILL.md","ichorCNA",[39,41,43,45,47],{"pre_analytics":40},"Tube type\u002Ftime-to-plasma in window; double-spin; low hemolysis; gDNA-contamination fragment check",{"after_consensus":42},"Unique-molecule coverage \u002F genome-equivalents recovered (NOT raw depth); duplication plateaued -- read both off the GroupReadsByUmi --family-size-histogram output",{"after_fragment":44},"Modal insert ~167bp (150-180); mononucleosome fraction >0.3",{"after_tf":46},"TF above the ~3% ichorCNA LoD to trust the value (below = below detection, not low burden)",{"after_mutation":48},"VarDict -f 0.005 is a reporting floor, NOT a detection threshold: confirm each call against a per-locus background-error model (PoN or smCounter2) before reporting; CHIP-subtracted against matched WBC","GPTomics\u002FbioSkills","main","2d1750717747097b602acf30bd09a7cbd4b4c6ea","awesome_skills\u002Fbiology\u002Fbio-workflows-liquid-biopsy-pipeline","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous","2026-11-09 11:45:36",{"source":8,"skillName":5,"commands":57,"prompt":67,"zipAvailable":25},[58,61,64],{"tool":59,"command":60},"npx","npx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill bio-workflows-liquid-biopsy-pipeline",{"tool":62,"command":63},"bunx","bunx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill bio-workflows-liquid-biopsy-pipeline",{"tool":65,"command":66},"pnpm","pnpm dlx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill bio-workflows-liquid-biopsy-pipeline","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-workflows-liquid-biopsy-pipeline\u002FSKILL.md",[69,75],{"path":70,"resourceType":71,"resourceTypeDesc":72,"sizeBytes":73,"sourceHtmlUrl":74},"examples\u002Fliquid_biopsy_pipeline.py","SCRIPT","脚本",8598,"https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-workflows-liquid-biopsy-pipeline\u002Fexamples\u002Fliquid_biopsy_pipeline.py",{"path":76,"resourceType":77,"resourceTypeDesc":78,"sizeBytes":79,"sourceHtmlUrl":80},"usage-guide.md","REFERENCE","参考文档",2786,"https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-workflows-liquid-biopsy-pipeline\u002Fusage-guide.md",[82,90,97,104,111,118],{"skillNo":83,"name":84,"nameZh":6,"description":85,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":86,"categories":87,"keywords":88,"publishedAt":89},"01M7FFFZPC2025JQQHPHDX154J","dei-assessment","Assess diversity, equity, and inclusion with metrics. TRIGGERS - Use when user needs help with dei-assessment related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fdei-assessment\u002FSKILL.md",[],[],"2026-11-16 10:05:34",{"skillNo":91,"name":92,"nameZh":6,"description":93,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":94,"categories":95,"keywords":96,"publishedAt":89},"01M7FFFWN8H4WA7A6Y398QKB6J","dei-action-plan","Create DEI action plans with goals and initiatives. TRIGGERS - Use when user needs help with dei-action-plan related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fdei-action-plan\u002FSKILL.md",[],[],{"skillNo":98,"name":99,"nameZh":6,"description":100,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":89},"01M7FFFSHTM0QZRVZ2NRJTJ0PY","culture-fit-assessment","Design culture add assessment processes. TRIGGERS - Use when user needs help with culture-fit-assessment related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fculture-fit-assessment\u002FSKILL.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":89},"01M7FFFMA12RKEQA38KHJSXXPS","cowork-to-code-bridge","Use an already-installed, independently verified cowork-to-code bridge to run narrowly approved actions on the user's own macOS, Linux, or WSL2 machine through a local file queue.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fcowork-to-code-bridge\u002FSKILL.md",[],[],{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":89},"01M7FFFH1F6M8YRGW5V487GH1K","counter-offer-response","Draft counter-offer responses for retention. TRIGGERS - Use when user needs help with counter-offer-response related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fcounter-offer-response\u002FSKILL.md",[],[],{"skillNo":119,"name":120,"nameZh":6,"description":121,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":122,"categories":123,"keywords":124,"publishedAt":89},"01M7FFFDGEHV8Y7J9N40HZXW52","continuous-feedback-system","Design continuous feedback systems. TRIGGERS - Use when user needs help with continuous-feedback-system related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fcontinuous-feedback-system\u002FSKILL.md",[],[]]