[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7EF64HAS1VQY2KHM81ZRT7M":3,"skill-same:01M7EF64HAS1VQY2KHM81ZRT7M":62},{"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":30,"commitSha":31,"skillDirectory":32,"forksCount":14,"repositoryHtmlUrl":33,"repositoryHomepageUrl":33,"ownerHtmlUrl":34,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":35,"install":36,"resources":49},"01M7EF64HAS1VQY2KHM81ZRT7M","bio-read-qc-contamination-screening",null,"Detects contamination in sequencing reads - cross-species (FastQ Screen, Kraken2), vector\u002FPhiX\u002Fadapter, rRNA, and same-species cross-sample\u002Findex-hopping and sample swaps (SNP fingerprints via verifyBamID2\u002FNGSCheckMate\u002Fsomalier). Use when suspecting cross-contamination, PDX host reads, microbial carry-over, or sample swaps, and to decide whether to report, filter, or align to a combined reference. For deep taxonomic profiling use metagenomics\u002Fkraken-classification.","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-read-qc-contamination-screening\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: bio-read-qc-contamination-screening\ndescription: Detects contamination in sequencing reads - cross-species (FastQ Screen, Kraken2), vector\u002FPhiX\u002Fadapter, rRNA, and same-species cross-sample\u002Findex-hopping and sample swaps (SNP fingerprints via verifyBamID2\u002FNGSCheckMate\u002Fsomalier). Use when suspecting cross-contamination, PDX host reads, microbial carry-over, or sample swaps, and to decide whether to report, filter, or align to a combined reference. For deep taxonomic profiling use metagenomics\u002Fkraken-classification.\ntool_type: cli\nprimary_tool: fastq_screen\nsource: \"https:\u002F\u002Fgithub.com\u002FGPTomics\u002FbioSkills\"\nsource_repository: \"GPTomics\u002FbioSkills\"\nsource_path: \"read-qc\u002Fcontamination-screening\u002FSKILL.md\"\nlicense: \"MIT\"\nimported_at: \"2026-09-20\"\n---\n\n## Version Compatibility\n\nReference examples tested with: FastQ Screen 0.15+, Bowtie2 2.5+, Kraken2 2.1+, BBTools 39.0+, MultiQC 1.21+\n\nBefore using code patterns, verify installed versions match. If versions differ:\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# Contamination Screening -- a species screen cannot see a same-species swap\n\nScreen reads against a genome panel (FastQ Screen \u002F Kraken2) for foreign ORGANISMS, and against SNP fingerprints for foreign or wrong INDIVIDUALS.\n\n**\"Check my reads for contamination\"** -> Map a subsample against multiple references and\u002For fingerprint sample identity to find foreign DNA and mislabels.\n- CLI: `fastq_screen --conf fastq_screen.conf sample.fastq.gz` (cross-species)\n- CLI: `verifyBamID2 --SVDPrefix resource --BamFile sample.bam` (same-species contamination)\n\nScope: this skill OWNS contamination detection and the report-vs-filter decision. Deep taxonomic profiling\u002Fabundance -> metagenomics\u002Fkraken-classification, metagenomics\u002Fmetaphlan-profiling. rRNA depletion as an RNA prep metric -> read-qc\u002Frnaseq-qc. OUT OF SCOPE: adapter removal (read-qc\u002Fadapter-trimming).\n\n## The Single Most Important Modern Insight\n\n1. **A SPECIES screen answers \"what ORGANISMS are here?\"; a SNP FINGERPRINT answers \"WHOSE DNA is this, and is it a mixture?\" -- and these are orthogonal.** A species screen is structurally BLIND to same-species cross-sample contamination and sample swaps: a human-A + human-B mixture, or a mislabeled human file, produces a perfectly clean single-species profile. Most pipelines run only a species screen and declare the data clean. Any human or single-species-cohort pipeline needs BOTH a taxonomic screen AND a SNP-fingerprint identity\u002Fcontamination check (verifyBamID2, NGSCheckMate, somalier, conpair).\n\n2. **Index hopping is the same-species contamination that lives inside one run, and unique dual indexing (UDI) is the only clean fix.** On patterned flowcells (HiSeq X\u002F4000, NovaSeq) ExAmp chemistry lets a free index adapter tag a fragment from another sample, spreading ~0.1-2% of reads into incorrect samples. Irrelevant for germline common variants, CATASTROPHIC for low-VAF work (ctDNA, single-cell, somatic) where hopped reads look like phantom low-frequency variants. Combinatorial indexing cannot detect it; UDI (a unique i7 AND i5 per sample) lets the demultiplexer drop impossible index pairs. UDI is effectively mandatory for ctDNA\u002Fplasma, single-cell, and low-input somatic.\n\n3. **Default to SCREEN-AND-REPORT, not filter -- removing reads biases composition.** A species screen is a QC gate; if a contaminant is low, aligning to the correct reference simply will not place the foreign reads. Filter only a specific, named contaminant (PhiX before assembly, adapters before alignment) with a precise k-mer remover (BBDuk `ref=phix`), never \"remove anything that hits the screen\" (that also discards conserved rRNA\u002Fmito reads that belong to the sample). For PDX, align to a COMBINED human+mouse reference and keep human-assigned reads, OR use a dedicated post-alignment classifier (XenofilteR benchmarks above Xenome for variant false-positive rate); both beat hard pre-filtering on one genome, which mis-assigns conserved-region reads.\n\nDeeper trap: reference-genome contamination corrupts the screen itself. A \"human\" hit can be bacterial sequence mis-deposited inside the human assembly (Conterminator found >2M contaminated GenBank entries). No `--confidence` setting fixes a wrong database; trust a deconned\u002Fcurated DB and treat surprising single-source hits as DB artifacts until ruled out.\n\n## The Contamination Taxonomy -- five classes, five fixes\n\n| Class | What it is | Detect with | Fix |\n|-------|-----------|-------------|-----|\n| Cross-species | Mouse in human PDX; bacteria in culture | FastQ Screen, Kraken2\u002FBracken, sourmash, Xenome\u002FXenofilteR | Combined-reference alignment; k-mer bin (report, do not blindly remove) |\n| Cross-sample \u002F index hopping | Same-species reads on the wrong sample (INVISIBLE to species screens) | verifyBamID2, NGSCheckMate, somalier, conpair | UDI at prep; drop impossible index pairs at demux |\n| Vector \u002F PhiX \u002F adapter | Spike-in, cloning vector, linkers | UniVec\u002FVecScreen; FastQ Screen adapter DB; BBDuk `ref=phix`\u002F`ref=adapters` | k-mer trim\u002Ffilter; upstream library QC |\n| rRNA over-representation | Library-prep failure, not contamination | SortMeRNA, ribodetector | Re-prep \u002F better depletion; filter only to recover depth |\n| Cell-line: mycoplasma \u002F misID | Mollicutes infection; HeLa cross-contamination | Kraken2 for Mollicutes; STR profiling for line identity | Clear culture or discard; STR-authenticate |\n\nA pipeline that runs FastQ Screen and stops has checked exactly one of five boxes.\n\n## Tool Taxonomy\n\n| Tool | Mechanism | When |\n|------|-----------|------|\n| FastQ Screen | Map a subsample to a genome panel; classify hit categories | Cross-species QC gate; the workhorse screen |\n| Kraken2 + Bracken | Exact-k-mer minimizer LCA classification; Bracken re-estimates abundance | Read-level taxonomy; many possible contaminants |\n| BBSplit \u002F BBDuk | k-mer binning \u002F named-contaminant k-mer removal | Decontamination of a NAMED contaminant (PhiX, adapters) |\n| Xenome \u002F XenofilteR | Classify reads human vs mouse (k-mer \u002F dual-alignment) | PDX host-graft disambiguation |\n| sourmash | MinHash\u002FFracMinHash containment sketches | Fast low-memory \"what is in here?\" screen |\n| verifyBamID2 | Per-sample within-species contamination from population SNP AFs | Same-species contamination level (FREEMIX) |\n| NGSCheckMate \u002F somalier | SNP-fingerprint identity \u002F relatedness | Sample swaps, tumor-normal pairing, longitudinal identity |\n| conpair | Tumor-normal concordance + independent contamination | Matched T\u002FN pairs |\n\n## Decision Tree by Scenario\n\n| Question | Use | Why |\n|----------|-----|-----|\n| Is a foreign ORGANISM present? | FastQ Screen or Kraken2 | Maps reads to species references |\n| Is this the right INDIVIDUAL \u002F one person? | NGSCheckMate \u002F somalier | SNP fingerprint, species-screen-blind |\n| What is the contamination LEVEL (human)? | verifyBamID2 (FREEMIX) | Estimates mixture fraction from SNP AFs |\n| Tumor-normal pair: matched and clean? | conpair | Concordance + per-sample contamination |\n| PDX host vs graft | Combined reference or a benchmarked classifier | XenofilteR > Xenome for SNV FP rate; both beat hard pre-filtering |\n| Remove a NAMED contaminant | BBDuk `ref=...` | Precise k-mer removal, not \"hits the screen\" |\n| Strip HUMAN reads before public deposition (non-human library) | hostile \u002F NCBI sra-human-scrubber (HRRT) | Deposition compliance, not a QC gate; a masked T2T reference avoids stripping conserved microbial regions |\n\nDefault when uncertain: FastQ Screen as the QC gate for organisms, PLUS a SNP-fingerprint check (somalier\u002FNGSCheckMate) for any human cohort.\n\n## FastQ Screen\n\nMaps a SUBSAMPLE (`--subset`, default 100000) with bowtie2 reporting >1 alignment, then classifies each read across the panel. Read the bar chart, not just \"% mapped\": contamination concentrates in `One_hit_one_genome` of an UNEXPECTED genome; homology (rRNA, mito, conserved loci) spreads into the `*_multiple_genomes` categories; high `Hit_no_genomes` means adapter dimer, a missing reference, or a novel organism (a diagnostic, not a verdict).\n\n```bash\n# Config: aligner binary + DATABASE lines (bowtie2 index prefixes)\ncat > fastq_screen.conf \u003C\u003C'EOF'\nBOWTIE2  \u002Fusr\u002Flocal\u002Fbin\u002Fbowtie2\nTHREADS  8\nDATABASE  Human  \u002Frefs\u002FGRCh38_bt2\u002FGRCh38\nDATABASE  Mouse  \u002Frefs\u002FGRCm39_bt2\u002FGRCm39\nDATABASE  Ecoli  \u002Frefs\u002FEcoli_bt2\u002FEcoli\nDATABASE  PhiX   \u002Frefs\u002Fphix_bt2\u002Fphix\nDATABASE  rRNA   \u002Frefs\u002FrRNA_bt2\u002FrRNA\nEOF\n\nfastq_screen --conf fastq_screen.conf --threads 8 --outdir screen\u002F *.fastq.gz\nmultiqc screen\u002F                      # MultiQC parses *_screen.txt across samples\n\n# Tag every read with a per-genome status, then extract a subset by pattern\nfastq_screen --conf fastq_screen.conf --tag --filter 10000 sample.fastq.gz   # maps only to genome 1\nfastq_screen --conf fastq_screen.conf --nohits sample.fastq.gz               # reads hitting nothing\n```\n\n`--filter` digits (one per genome, config order): 0=no map, 1=unique, 2=multi, 3=maps, 4=pass 0 or 1, 5=pass 0 or 2, -=ignore. `--subset 0` screens the whole file; `--bisulfite` uses Bismark.\n\n## Kraken2 + Bracken (read-level taxonomy)\n\nDefault `--confidence 0.0` over-reports a long tail of spurious low-abundance species (a few shared k-mers suffice); raise to 0.05-0.1 and keep `--minimum-hit-groups 2` (or 3 for custom DBs). Kraken2 gives CLASSIFICATION; Bracken redistributes higher-rank reads to species for ABUNDANCE.\n\n```bash\nkraken2 --db \u002Fdb\u002Fk2_standard --threads 8 --confidence 0.1 --paired \\\n        --report sample.kreport --use-names R1.fq.gz R2.fq.gz > sample.kraken\nbracken -d \u002Fdb\u002Fk2_standard -i sample.kreport -o sample.bracken -r 150 -l S\n```\n\n## Same-species: SNP fingerprints and index hopping\n\n```bash\n# verifyBamID2: FREEMIX = contamination fraction (action threshold ~0.02);\n# FREEMIX~0 with CHIPMIX~1 indicates a SWAP, not contamination.\n# --SVDPrefix points to the panel resource that ships with verifyBamID2 (resource\u002F1000g.phase3...).\nverifyBamID2 --SVDPrefix \u002Fres\u002F1000g.phase3.100k.b38.vcf.gz.dat --BamFile sample.bam --Reference ref.fa\n\n# somalier: extract genome sketches, then relate to find swaps \u002F identity across a cohort.\n# --sites = somalier's released sites.\u003Cbuild>.vcf.gz (github releases), not a custom panel.\nsomalier extract -d sites\u002F --sites sites.hg38.vcf.gz -f ref.fa sample.bam\nsomalier relate sites\u002F*.somalier            # off-diagonal identity flags swaps\n\n# conpair (tumor-normal): concordance + independent per-sample contamination\n```\n\nIndex hopping is mitigated at demultiplexing with UDI (drop impossible i7,i5 pairs); residual contamination is then quantified by the SNP-fingerprint tools above.\n\n## Common Errors\n\n| Symptom | Cause | Solution |\n|---------|-------|----------|\n| \"Single species, data is clean\" but a swap is suspected | Species screen is blind to same-species swaps | Run somalier \u002F NGSCheckMate on SNP fingerprints |\n| Phantom low-VAF variants in ctDNA\u002Fsingle-cell | Index hopping on patterned flowcell | Use UDI; quantify residual with verifyBamID2\u002Fconpair |\n| Kraken2 reports dozens of trace species | Default confidence 0.0 over-reports | Raise `--confidence` to 0.05-0.1; raise hit-groups |\n| A \"human\" Kraken hit on a microbial isolate | Reference\u002FDB contamination | Use a deconned DB; treat as artifact until confirmed |\n| Filtering \"contaminant\" reads skews composition | Removed conserved rRNA\u002Fmito too | Remove a NAMED contaminant with BBDuk, not \"hits the screen\" |\n| PDX human counts look biased | Hard pre-filtering of ambiguous reads | Align to combined human+mouse reference instead |\n\n## References\n\nWingett SW, Andrews S. 2018. FastQ Screen: a tool for multi-genome mapping and quality control. F1000Research 7:1338.\nWood DE, Lu J, Langmead B. 2019. Improved metagenomic analysis with Kraken 2. Genome Biology 20:257.\nLu J, Breitwieser FP, Thielen P, Salzberg SL. 2017. Bracken: estimating species abundance in metagenomics data. PeerJ Computer Science 3:e104.\nSteinegger M, Salzberg SL. 2020. Terminating contamination: large-scale search identifies more than 2,000,000 contaminated entries in GenBank. Genome Biology 21:115.\nConway T, Wazny J, Bromage A, et al. 2012. Xenome - a tool for classifying reads from xenograft samples. Bioinformatics 28(12):i172-i178.\nCostello M, Fleharty M, Abreu J, et al. 2018. Characterization and remediation of sample index swaps by non-redundant dual indexing. BMC Genomics 19:332.\nZhang F, Flickinger M, Taliun SAG, et al. 2020. Ancestry-agnostic estimation of DNA sample contamination from sequence reads. Genome Research 30(2):185-194.\nLee S, Lee S, Ouellette S, Park WY, Lee EA, Park PJ. 2017. NGSCheckMate: software for validating sample identity in next-generation sequencing studies within and across data types. Nucleic Acids Research 45(11):e103.\nPedersen BS, Bhetariya PJ, Brown J, et al. 2020. Somalier: rapid relatedness estimation for cancer and germline studies using efficient genome sketches. Genome Medicine 12:62.\n\n## Related Skills\n\nread-qc\u002Fquality-reports - Bimodal GC and overrepresented sequences flag contamination\nread-qc\u002Fadapter-trimming - Remove adapter contamination\nread-qc\u002Frnaseq-qc - rRNA fraction as a prep-efficiency metric\nmetagenomics\u002Fkraken-classification - Deeper taxonomic classification and profiling\nvariant-calling\u002Fjoint-calling - Where SNP-fingerprint sample swaps do the most damage\n","## Version Compatibility\n\nReference examples tested with: FastQ Screen 0.15+, Bowtie2 2.5+, Kraken2 2.1+, BBTools 39.0+, MultiQC 1.21+\n\nBefore using code patterns, verify installed versions match. If versions differ:\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# Contamination Screening -- a species screen cannot see a same-species swap\n\nScreen reads against a genome panel (FastQ Screen \u002F Kraken2) for foreign ORGANISMS, and against SNP fingerprints for foreign or wrong INDIVIDUALS.\n\n**\"Check my reads for contamination\"** -> Map a subsample against multiple references and\u002For fingerprint sample identity to find foreign DNA and mislabels.\n- CLI: `fastq_screen --conf fastq_screen.conf sample.fastq.gz` (cross-species)\n- CLI: `verifyBamID2 --SVDPrefix resource --BamFile sample.bam` (same-species contamination)\n\nScope: this skill OWNS contamination detection and the report-vs-filter decision. Deep taxonomic profiling\u002Fabundance -> metagenomics\u002Fkraken-classification, metagenomics\u002Fmetaphlan-profiling. rRNA depletion as an RNA prep metric -> read-qc\u002Frnaseq-qc. OUT OF SCOPE: adapter removal (read-qc\u002Fadapter-trimming).\n\n## The Single Most Important Modern Insight\n\n1. **A SPECIES screen answers \"what ORGANISMS are here?\"; a SNP FINGERPRINT answers \"WHOSE DNA is this, and is it a mixture?\" -- and these are orthogonal.** A species screen is structurally BLIND to same-species cross-sample contamination and sample swaps: a human-A + human-B mixture, or a mislabeled human file, produces a perfectly clean single-species profile. Most pipelines run only a species screen and declare the data clean. Any human or single-species-cohort pipeline needs BOTH a taxonomic screen AND a SNP-fingerprint identity\u002Fcontamination check (verifyBamID2, NGSCheckMate, somalier, conpair).\n\n2. **Index hopping is the same-species contamination that lives inside one run, and unique dual indexing (UDI) is the only clean fix.** On patterned flowcells (HiSeq X\u002F4000, NovaSeq) ExAmp chemistry lets a free index adapter tag a fragment from another sample, spreading ~0.1-2% of reads into incorrect samples. Irrelevant for germline common variants, CATASTROPHIC for low-VAF work (ctDNA, single-cell, somatic) where hopped reads look like phantom low-frequency variants. Combinatorial indexing cannot detect it; UDI (a unique i7 AND i5 per sample) lets the demultiplexer drop impossible index pairs. UDI is effectively mandatory for ctDNA\u002Fplasma, single-cell, and low-input somatic.\n\n3. **Default to SCREEN-AND-REPORT, not filter -- removing reads biases composition.** A species screen is a QC gate; if a contaminant is low, aligning to the correct reference simply will not place the foreign reads. Filter only a specific, named contaminant (PhiX before assembly, adapters before alignment) with a precise k-mer remover (BBDuk `ref=phix`), never \"remove anything that hits the screen\" (that also discards conserved rRNA\u002Fmito reads that belong to the sample). For PDX, align to a COMBINED human+mouse reference and keep human-assigned reads, OR use a dedicated post-alignment classifier (XenofilteR benchmarks above Xenome for variant false-positive rate); both beat hard pre-filtering on one genome, which mis-assigns conserved-region reads.\n\nDeeper trap: reference-genome contamination corrupts the screen itself. A \"human\" hit can be bacterial sequence mis-deposited inside the human assembly (Conterminator found >2M contaminated GenBank entries). No `--confidence` setting fixes a wrong database; trust a deconned\u002Fcurated DB and treat surprising single-source hits as DB artifacts until ruled out.\n\n## The Contamination Taxonomy -- five classes, five fixes\n\n| Class | What it is | Detect with | Fix |\n|-------|-----------|-------------|-----|\n| Cross-species | Mouse in human PDX; bacteria in culture | FastQ Screen, Kraken2\u002FBracken, sourmash, Xenome\u002FXenofilteR | Combined-reference alignment; k-mer bin (report, do not blindly remove) |\n| Cross-sample \u002F index hopping | Same-species reads on the wrong sample (INVISIBLE to species screens) | verifyBamID2, NGSCheckMate, somalier, conpair | UDI at prep; drop impossible index pairs at demux |\n| Vector \u002F PhiX \u002F adapter | Spike-in, cloning vector, linkers | UniVec\u002FVecScreen; FastQ Screen adapter DB; BBDuk `ref=phix`\u002F`ref=adapters` | k-mer trim\u002Ffilter; upstream library QC |\n| rRNA over-representation | Library-prep failure, not contamination | SortMeRNA, ribodetector | Re-prep \u002F better depletion; filter only to recover depth |\n| Cell-line: mycoplasma \u002F misID | Mollicutes infection; HeLa cross-contamination | Kraken2 for Mollicutes; STR profiling for line identity | Clear culture or discard; STR-authenticate |\n\nA pipeline that runs FastQ Screen and stops has checked exactly one of five boxes.\n\n## Tool Taxonomy\n\n| Tool | Mechanism | When |\n|------|-----------|------|\n| FastQ Screen | Map a subsample to a genome panel; classify hit categories | Cross-species QC gate; the workhorse screen |\n| Kraken2 + Bracken | Exact-k-mer minimizer LCA classification; Bracken re-estimates abundance | Read-level taxonomy; many possible contaminants |\n| BBSplit \u002F BBDuk | k-mer binning \u002F named-contaminant k-mer removal | Decontamination of a NAMED contaminant (PhiX, adapters) |\n| Xenome \u002F XenofilteR | Classify reads human vs mouse (k-mer \u002F dual-alignment) | PDX host-graft disambiguation |\n| sourmash | MinHash\u002FFracMinHash containment sketches | Fast low-memory \"what is in here?\" screen |\n| verifyBamID2 | Per-sample within-species contamination from population SNP AFs | Same-species contamination level (FREEMIX) |\n| NGSCheckMate \u002F somalier | SNP-fingerprint identity \u002F relatedness | Sample swaps, tumor-normal pairing, longitudinal identity |\n| conpair | Tumor-normal concordance + independent contamination | Matched T\u002FN pairs |\n\n## Decision Tree by Scenario\n\n| Question | Use | Why |\n|----------|-----|-----|\n| Is a foreign ORGANISM present? | FastQ Screen or Kraken2 | Maps reads to species references |\n| Is this the right INDIVIDUAL \u002F one person? | NGSCheckMate \u002F somalier | SNP fingerprint, species-screen-blind |\n| What is the contamination LEVEL (human)? | verifyBamID2 (FREEMIX) | Estimates mixture fraction from SNP AFs |\n| Tumor-normal pair: matched and clean? | conpair | Concordance + per-sample contamination |\n| PDX host vs graft | Combined reference or a benchmarked classifier | XenofilteR > Xenome for SNV FP rate; both beat hard pre-filtering |\n| Remove a NAMED contaminant | BBDuk `ref=...` | Precise k-mer removal, not \"hits the screen\" |\n| Strip HUMAN reads before public deposition (non-human library) | hostile \u002F NCBI sra-human-scrubber (HRRT) | Deposition compliance, not a QC gate; a masked T2T reference avoids stripping conserved microbial regions |\n\nDefault when uncertain: FastQ Screen as the QC gate for organisms, PLUS a SNP-fingerprint check (somalier\u002FNGSCheckMate) for any human cohort.\n\n## FastQ Screen\n\nMaps a SUBSAMPLE (`--subset`, default 100000) with bowtie2 reporting >1 alignment, then classifies each read across the panel. Read the bar chart, not just \"% mapped\": contamination concentrates in `One_hit_one_genome` of an UNEXPECTED genome; homology (rRNA, mito, conserved loci) spreads into the `*_multiple_genomes` categories; high `Hit_no_genomes` means adapter dimer, a missing reference, or a novel organism (a diagnostic, not a verdict).\n\n```bash\n# Config: aligner binary + DATABASE lines (bowtie2 index prefixes)\ncat > fastq_screen.conf \u003C\u003C'EOF'\nBOWTIE2  \u002Fusr\u002Flocal\u002Fbin\u002Fbowtie2\nTHREADS  8\nDATABASE  Human  \u002Frefs\u002FGRCh38_bt2\u002FGRCh38\nDATABASE  Mouse  \u002Frefs\u002FGRCm39_bt2\u002FGRCm39\nDATABASE  Ecoli  \u002Frefs\u002FEcoli_bt2\u002FEcoli\nDATABASE  PhiX   \u002Frefs\u002Fphix_bt2\u002Fphix\nDATABASE  rRNA   \u002Frefs\u002FrRNA_bt2\u002FrRNA\nEOF\n\nfastq_screen --conf fastq_screen.conf --threads 8 --outdir screen\u002F *.fastq.gz\nmultiqc screen\u002F                      # MultiQC parses *_screen.txt across samples\n\n# Tag every read with a per-genome status, then extract a subset by pattern\nfastq_screen --conf fastq_screen.conf --tag --filter 10000 sample.fastq.gz   # maps only to genome 1\nfastq_screen --conf fastq_screen.conf --nohits sample.fastq.gz               # reads hitting nothing\n```\n\n`--filter` digits (one per genome, config order): 0=no map, 1=unique, 2=multi, 3=maps, 4=pass 0 or 1, 5=pass 0 or 2, -=ignore. `--subset 0` screens the whole file; `--bisulfite` uses Bismark.\n\n## Kraken2 + Bracken (read-level taxonomy)\n\nDefault `--confidence 0.0` over-reports a long tail of spurious low-abundance species (a few shared k-mers suffice); raise to 0.05-0.1 and keep `--minimum-hit-groups 2` (or 3 for custom DBs). Kraken2 gives CLASSIFICATION; Bracken redistributes higher-rank reads to species for ABUNDANCE.\n\n```bash\nkraken2 --db \u002Fdb\u002Fk2_standard --threads 8 --confidence 0.1 --paired \\\n        --report sample.kreport --use-names R1.fq.gz R2.fq.gz > sample.kraken\nbracken -d \u002Fdb\u002Fk2_standard -i sample.kreport -o sample.bracken -r 150 -l S\n```\n\n## Same-species: SNP fingerprints and index hopping\n\n```bash\n# verifyBamID2: FREEMIX = contamination fraction (action threshold ~0.02);\n# FREEMIX~0 with CHIPMIX~1 indicates a SWAP, not contamination.\n# --SVDPrefix points to the panel resource that ships with verifyBamID2 (resource\u002F1000g.phase3...).\nverifyBamID2 --SVDPrefix \u002Fres\u002F1000g.phase3.100k.b38.vcf.gz.dat --BamFile sample.bam --Reference ref.fa\n\n# somalier: extract genome sketches, then relate to find swaps \u002F identity across a cohort.\n# --sites = somalier's released sites.\u003Cbuild>.vcf.gz (github releases), not a custom panel.\nsomalier extract -d sites\u002F --sites sites.hg38.vcf.gz -f ref.fa sample.bam\nsomalier relate sites\u002F*.somalier            # off-diagonal identity flags swaps\n\n# conpair (tumor-normal): concordance + independent per-sample contamination\n```\n\nIndex hopping is mitigated at demultiplexing with UDI (drop impossible i7,i5 pairs); residual contamination is then quantified by the SNP-fingerprint tools above.\n\n## Common Errors\n\n| Symptom | Cause | Solution |\n|---------|-------|----------|\n| \"Single species, data is clean\" but a swap is suspected | Species screen is blind to same-species swaps | Run somalier \u002F NGSCheckMate on SNP fingerprints |\n| Phantom low-VAF variants in ctDNA\u002Fsingle-cell | Index hopping on patterned flowcell | Use UDI; quantify residual with verifyBamID2\u002Fconpair |\n| Kraken2 reports dozens of trace species | Default confidence 0.0 over-reports | Raise `--confidence` to 0.05-0.1; raise hit-groups |\n| A \"human\" Kraken hit on a microbial isolate | Reference\u002FDB contamination | Use a deconned DB; treat as artifact until confirmed |\n| Filtering \"contaminant\" reads skews composition | Removed conserved rRNA\u002Fmito too | Remove a NAMED contaminant with BBDuk, not \"hits the screen\" |\n| PDX human counts look biased | Hard pre-filtering of ambiguous reads | Align to combined human+mouse reference instead |\n\n## References\n\nWingett SW, Andrews S. 2018. FastQ Screen: a tool for multi-genome mapping and quality control. F1000Research 7:1338.\nWood DE, Lu J, Langmead B. 2019. Improved metagenomic analysis with Kraken 2. Genome Biology 20:257.\nLu J, Breitwieser FP, Thielen P, Salzberg SL. 2017. Bracken: estimating species abundance in metagenomics data. PeerJ Computer Science 3:e104.\nSteinegger M, Salzberg SL. 2020. Terminating contamination: large-scale search identifies more than 2,000,000 contaminated entries in GenBank. Genome Biology 21:115.\nConway T, Wazny J, Bromage A, et al. 2012. Xenome - a tool for classifying reads from xenograft samples. Bioinformatics 28(12):i172-i178.\nCostello M, Fleharty M, Abreu J, et al. 2018. Characterization and remediation of sample index swaps by non-redundant dual indexing. BMC Genomics 19:332.\nZhang F, Flickinger M, Taliun SAG, et al. 2020. Ancestry-agnostic estimation of DNA sample contamination from sequence reads. Genome Research 30(2):185-194.\nLee S, Lee S, Ouellette S, Park WY, Lee EA, Park PJ. 2017. NGSCheckMate: software for validating sample identity in next-generation sequencing studies within and across data types. Nucleic Acids Research 45(11):e103.\nPedersen BS, Bhetariya PJ, Brown J, et al. 2020. Somalier: rapid relatedness estimation for cancer and germline studies using efficient genome sketches. Genome Medicine 12:62.\n\n## Related Skills\n\nread-qc\u002Fquality-reports - Bimodal GC and overrepresented sequences flag contamination\nread-qc\u002Fadapter-trimming - Remove adapter contamination\nread-qc\u002Frnaseq-qc - rRNA fraction as a prep-efficiency metric\nmetagenomics\u002Fkraken-classification - Deeper taxonomic classification and profiling\nvariant-calling\u002Fjoint-calling - Where SNP-fingerprint sample swaps do the most damage\n",{"name":5,"source":24,"license":15,"tool_type":25,"description":7,"imported_at":26,"source_path":27,"primary_tool":28,"source_repository":29},"https:\u002F\u002Fgithub.com\u002FGPTomics\u002FbioSkills","cli","2026-09-20","read-qc\u002Fcontamination-screening\u002FSKILL.md","fastq_screen","GPTomics\u002FbioSkills","main","2d1750717747097b602acf30bd09a7cbd4b4c6ea","awesome_skills\u002Fbiology\u002Fbio-read-qc-contamination-screening","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":37,"prompt":47,"zipAvailable":48},[38,41,44],{"tool":39,"command":40},"npx","npx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill bio-read-qc-contamination-screening",{"tool":42,"command":43},"bunx","bunx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill bio-read-qc-contamination-screening",{"tool":45,"command":46},"pnpm","pnpm dlx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill bio-read-qc-contamination-screening","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-read-qc-contamination-screening\u002FSKILL.md",true,[50,56],{"path":51,"resourceType":52,"resourceTypeDesc":53,"sizeBytes":54,"sourceHtmlUrl":55},"examples\u002Fscreen_samples.sh","SCRIPT","脚本",1335,"https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-read-qc-contamination-screening\u002Fexamples\u002Fscreen_samples.sh",{"path":57,"resourceType":58,"resourceTypeDesc":59,"sizeBytes":60,"sourceHtmlUrl":61},"usage-guide.md","REFERENCE","参考文档",3924,"https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fbiology\u002Fbio-read-qc-contamination-screening\u002Fusage-guide.md",[63,71,78,85,92,99],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":70},"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":72,"name":73,"nameZh":6,"description":74,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":75,"categories":76,"keywords":77,"publishedAt":70},"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":79,"name":80,"nameZh":6,"description":81,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":82,"categories":83,"keywords":84,"publishedAt":70},"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":86,"name":87,"nameZh":6,"description":88,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":70},"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":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":70},"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":100,"name":101,"nameZh":6,"description":102,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":103,"categories":104,"keywords":105,"publishedAt":70},"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",[],[]]