bio-chromosome

name	bio-chromosome
description	Chromosome- and assembly-scale genome analysis with Seqeron (MCP tools OR the C# API). Use for karyotype & ploidy / aneuploidy (trisomy, monosomy) from depth or chromosome descriptors, centromere / telomere / heterochromatin and G-band prediction, arm-ratio classification (metacentric/acrocentric), synteny blocks & large structural rearrangements (inversion/translocation/ deletion/duplication), assembly QC (N50/L50, auN, Nx curve, contiguity, gaps, contig extraction, completeness/BUSCO-like, assembly comparison), and chromosome-scale composition/GC-skew profiling. Triggers: "analyze this karyotype", "detect aneuploidy / ploidy", "find the centromere / telomeres", "N50 of this assembly", "assembly stats", "find synteny blocks", "detect rearrangements", "GC-skew of this chromosome", "predict the replication origin", "sliding-window GC / composition". Server: chromosome (+ analysis for GC-skew/origin).
allowed-tools	Read, Bash, Grep, Glob

bio-chromosome — karyotype, centromere/telomere, assembly QC, synteny/SV, composition

Routing + orchestration skill for the Chromosome server (32 tools: ChromosomeAnalyzer.* + GenomeAssemblyAnalyzer.*). It picks the right tool for a chromosome- or assembly-scale question and gives a dual-mode recipe (MCP tool calls and the equivalent Seqeron.Genomics C# Method ID).

• Rigor is delegated. Parse-with-a-tool, envelope, provenance, cross-check, units / 0-based (or here often 1-based inclusive) coordinates, and the alpha / not-for-clinical caveat are owned by bio-rigor — it applies by default; do not restate.
• Don't know a tool name? Use seqeron-discovery (python3 scripts/skills/find-tool.py <kw> --server chromosome) — never guess.
• Point, don't duplicate. Full I/O schemas live in docs/mcp/tools/chromosome/*.md; algorithm invariants in docs/algorithms/{Chromosome_Analysis,Assembly,Extended_GC_Skew_Analysis}/*.md.

Scope boundary — this server vs neighbours

• GC-skew, cumulative/windowed GC-skew, replication-origin prediction live on the analysis server, not chromosome: gc_skew / cumulative_gc_skew / windowed_gc_skew / predict_replication_origin (GcSkewCalculator.*) — route via bio-annotation. Composition here = G-bands (predict_g_bands), local GC/complexity (local_quality), suspicious windows (find_suspicious_regions).
• Variant-level read-evidence SV / CNV (single breakpoints, per-locus CNV calls) → seqeron-structural-variants; VEP/ACMG variant effects → bio-annotation. This skill = CHROMOSOME-SCALE structure: synteny-based rearrangements and depth-based whole-chromosome aneuploidy/ploidy. Gene-order-permutation rearrangements, reversal distance, ANI/orthologs/RBH and dot-plots → seqeron-comparative-genomics.

Decision guide — which tool for which question

Question	Tool ([MCP] / Method ID)
Karyotype from chromosome descriptors → aneuploidy flags	analyze_karyotype / ChromosomeAnalyzer.AnalyzeKaryotype
Whole-chromosome aneuploidy from read depth	identify_whole_chromosome_aneuploidy / …IdentifyWholeChromosomeAneuploidy
Per-bin copy number / aneuploidy from depth	detect_aneuploidy / ChromosomeAnalyzer.DetectAneuploidy
Overall ploidy from normalized depth	detect_ploidy / ChromosomeAnalyzer.DetectPloidy
Locate centromere (alpha-satellite scan)	analyze_centromere / ChromosomeAnalyzer.AnalyzeCentromere
Arm ratio from centromere position	arm_ratio / ChromosomeAnalyzer.CalculateArmRatio
Classify centromere position (meta/acro…)	classify_chromosome_by_arm_ratio / …ClassifyChromosomeByArmRatio
Telomere tract presence / critical shortening	analyze_telomeres / ChromosomeAnalyzer.AnalyzeTelomeres
Telomere length from qPCR T/S ratio	estimate_telomere_length_from_ts_ratio / …EstimateTelomereLengthFromTSRatio
Cell divisions from telomere shortening	estimate_cell_divisions_from_telomere_length / …EstimateCellDivisionsFromTelomereLength
Heterochromatin regions (repeat-content windows)	find_heterochromatin_regions / …FindHeterochromatinRegions
GC-based cytogenetic G-bands	predict_g_bands / ChromosomeAnalyzer.PredictGBands
Synteny blocks from ortholog pairs	find_synteny_blocks / ChromosomeAnalyzer.FindSyntenyBlocks
Rearrangements (inv/transloc/del/dup) from blocks	detect_rearrangements / ChromosomeAnalyzer.DetectRearrangements
Synteny blocks between two assemblies (k-mer)	find_syntenic_blocks_assemblies / GenomeAssemblyAnalyzer.FindSyntenicBlocks
Assembly stats (N50/L50, GC, gaps)	assembly_statistics / GenomeAssemblyAnalyzer.CalculateStatistics
Nx/Lx at a threshold · full Nx curve · auN	nx_statistics · nx_curve · au_n (GenomeAssemblyAnalyzer.CalculateNx / CalculateNxCurve / CalculateAuN)
Length distribution of contigs/scaffolds	length_distribution / …CalculateLengthDistribution
Find / distribution of N-gaps	find_gaps · gap_distribution (…FindGaps / …AnalyzeGapDistribution)
Split scaffolds → contigs	extract_contigs · analyze_scaffolds (…ExtractContigs / …AnalyzeScaffolds)
Completeness: marker-gene (BUSCO-like) · k-mer spectrum	assess_completeness · estimate_completeness_from_kmers
Local per-window quality (GC/N/complexity)	local_quality / GenomeAssemblyAnalyzer.CalculateLocalQuality
Suspicious windows (GC/complexity outliers)	find_suspicious_regions / …FindSuspiciousRegions
Tandem repeats · repetitive regions · repeat content	find_tandem_repeats · find_repetitive_regions · repeat_content
Compare two assemblies (shared k-mers)	compare_assemblies / GenomeAssemblyAnalyzer.CompareAssemblies

Canonical dual-mode pipelines

(a) Chromosome sequence → centromere → arm ratio → classify

1. [MCP] analyze_centromere(chromosomeName, sequence, windowSize?=100000, minAlphaSatelliteContent?=0.3) → centromere start/end (null if none), type.
2. [MCP] arm_ratio(centromerePosition, chromosomeLength) → p, q, armRatio.
3. [MCP] classify_chromosome_by_arm_ratio(armRatio) → Metacentric / Submetacentric / Subtelocentric / Acrocentric / Telocentric.

• [C# API] ChromosomeAnalyzer.AnalyzeCentromere(...) → .CalculateArmRatio(pos,len) → .ClassifyChromosomeByArmRatio(r).

Provenance
1) analyze_centromere(name,seq,windowSize=100000,minAlphaSatelliteContent=0.3) → start,end,type
2) arm_ratio(centromerePosition=start,chromosomeLength=len) → p,q,armRatio
3) classify_chromosome_by_arm_ratio(armRatio) → type
Cross-check: analyze_centromere's own `type` (Levan) should agree with step 3's classification.
Envelope: none of these guarded. Caveat: alpha — validate before decision use.

(b) Read depth → whole-chromosome aneuploidy / ploidy

1. [MCP] detect_aneuploidy(depthData=[{chromosome,position,depth}], medianDepth, binSize?=1000000) → per-bin copyNumber, logRatio, confidence.
2. [MCP] identify_whole_chromosome_aneuploidy(depthData, medianDepth, binSize?, minFraction?) → per-chromosome dominant CN ≠ 2 as ISCN aneuploidy.
3. [MCP] (genome-wide) detect_ploidy(normalizedDepths, expectedDiploidDepth) → overall ploidy.

• [C# API] ChromosomeAnalyzer.DetectAneuploidy(...) → .IdentifyWholeChromosomeAneuploidy(...) → .DetectPloidy(...).
• Descriptor-only alternative (no depth): analyze_karyotype(chromosomes=[{name,length,isSexChromosome}], expectedPloidyLevel?=2) → abnormalities.

(c) Ortholog pairs → synteny blocks → large structural rearrangements

1. [MCP] find_synteny_blocks(orthologPairs=[{chr1,start1,end1,gene1,chr2,start2,end2,gene2}], minGenes?=3, maxGap?=10) → collinear blocks + strand.
2. [MCP] detect_rearrangements(syntenyBlocks) → inversion / translocation / deletion / duplication events + breakpoints.

• [C# API] ChromosomeAnalyzer.FindSyntenyBlocks(...) → .DetectRearrangements(blocks).
• Assembly-vs-assembly (no genes): find_syntenic_blocks_assemblies(assembly1, assembly2, minBlockSize?) → k-mer-anchored blocks with orientation flags.
• Variant-level breakpoints/CNV (read evidence) → cross-link seqeron-structural-variants.

(d) Assembly QC — contiguity, gaps, completeness

1. [MCP] assembly_statistics(sequences=[{id,sequence}]) → N50/L50, N90/L90, GC, gap counts, largest/mean/median.
2. [MCP] au_n(lengths) → threshold-free contiguity; nx_curve(sortedLengths, totalLength, thresholds?) → full N10…N90 curve.
3. [MCP] find_gaps(sequences, minGapLength?=1) → [start,end] inclusive per N-run; gap_distribution(gaps) → summary.
4. [MCP] completeness: assess_completeness(markerGenes, assembly) (BUSCO-like) or estimate_completeness_from_kmers(kmerSpectrum).

• [C# API] GenomeAssemblyAnalyzer.CalculateStatistics(...) · .CalculateAuN(...) · .CalculateNxCurve(...) · .FindGaps(...) · .AnalyzeGapDistribution(...) · .AssessCompleteness(...).
• Note: nx_statistics/nx_curve expect descending-sorted lengths + precomputed totalLength; feed them from length_distribution/your own sort, not raw sequences.

(e) Chromosome-scale composition profile

1. [MCP] predict_g_bands(sequence, bandSize?, darkBandGcThreshold?, lightBandGcThreshold?) → gpos100/gpos50/gneg bands by windowed GC.
2. [MCP] local_quality(sequences, windowSize?) → per-window GC / N-count / linguistic complexity (step = windowSize/2).
3. [MCP] find_heterochromatin_regions(sequence, windowSize?, minRepeatContent?) → merged repeat-rich regions, classed Telomeric/Centromeric/Constitutive.

• [C# API] ChromosomeAnalyzer.PredictGBands(...) · GenomeAssemblyAnalyzer.CalculateLocalQuality(...) · ChromosomeAnalyzer.FindHeterochromatinRegions(...).
• GC-skew / cumulative skew / replication origin are on the analysis server → use bio-annotation: gc_skew, cumulative_gc_skew, windowed_gc_skew, predict_replication_origin (GcSkewCalculator.*).

(f) Telomere biology

1. [MCP] analyze_telomeres(chromosomeName, sequence, telomereRepeat?=TTAGGG, searchLength?, minTelomereLength?, criticalLength?) → present? / length / purity / critically-short.
2. [MCP] estimate_telomere_length_from_ts_ratio(tsRatio, referenceRatio, referenceLength) → length from qPCR.
3. [MCP] estimate_cell_divisions_from_telomere_length(birthLength, currentLength, lossPerDivision) → replicative age proxy.

• [C# API] ChromosomeAnalyzer.AnalyzeTelomeres(...) · .EstimateTelomereLengthFromTSRatio(...) · .EstimateCellDivisionsFromTelomereLength(...).

Envelope — STOP rule (guarded unit in scope)

• CHROM-CENT-001 (AssignSuprachromosomalFamily, output Sf1OrSf2Dimeric) is Permissive-only (blocked in Strict & Moderate; default is Moderate). Resolving SF1 vs SF2 needs an SF-resolved consensus-monomer reference the library does not ship. If a task asks to assign suprachromosomal family / distinguish SF1 vs SF2, STOP and report the envelope — do not force a result. The general centromere tools above (analyze_centromere, etc.) are not guarded.

End-to-end grounded example (extends docs/mcp/README.md)

Task. Given a draft assembly (a few {id,sequence} scaffolds), QC it and interpret one chromosome: (1) contiguity + gaps, (2) split its largest scaffold into contigs, (3) locate that chromosome's centromere and classify it, (4) profile heterochromatin, (5) corroborate contiguity threshold-free.

Tool / Method ID chain (MCP names; C# path in parentheses):

1. assembly_statistics(sequences) → n50, l50, gcContent, gapCount. (GenomeAssemblyAnalyzer.CalculateStatistics)
2. find_gaps(sequences, minGapLength=10) then gap_distribution(gaps) → gap classes. (…FindGaps / …AnalyzeGapDistribution)
3. extract_contigs(scaffolds, minContigLength=500) → {id}_contig{n} sequences over gap-free runs. (…ExtractContigs)
4. analyze_centromere(chromosomeName, sequence) → start/end/type; then arm_ratio(start, length) → armRatio; then classify_chromosome_by_arm_ratio(armRatio). (ChromosomeAnalyzer.AnalyzeCentromere → .CalculateArmRatio → .ClassifyChromosomeByArmRatio)
5. find_heterochromatin_regions(sequence) → Centromeric region should overlap the centromere from step 4 (independent repeat-content path). (…FindHeterochromatinRegions)
6. Cross-check contiguity: au_n(lengths) — threshold-free auN should track N50 direction. (GenomeAssemblyAnalyzer.CalculateAuN)

Provenance
1) assembly_statistics(sequences) → n50,l50,gcContent,gapCount
2) find_gaps(sequences,minGapLength=10)+gap_distribution → gap classes ([start,end] inclusive)
3) extract_contigs(scaffolds,minContigLength=500) → contigs
4) analyze_centromere→arm_ratio→classify_chromosome_by_arm_ratio → centromere type
5) find_heterochromatin_regions → Centromeric region overlaps step-4 centromere (cross-check)
6) au_n(lengths) → threshold-free contiguity consistent with N50
Envelope: none of the above guarded (CHROM-CENT-001 applies only to SF1/SF2 assignment — not invoked here).
Caveat: alpha software; not for clinical use — independently validate before relying on any call.

Reference

• Full domain tool index (all 32, generated — do NOT hand-edit): _generated/tools.md (produced by scripts/skills/gen-catalog.py; if absent, run seqeron-discovery).
• Fuller recipes + parameter guidance: reference/pipelines.md
• Tool map (all 32 by group, one-liners + Method ID + doc): reference/tool-map.md
• Algorithm background (invariants/formulas — link, don't copy): Karyotype_Analysis.md · Centromere_Analysis.md · Telomere_Analysis.md · Synteny_Analysis.md · Aneuploidy_Detection.md · Assembly_Statistics.md · Comprehensive_GC_Analysis.md
• Cross-cutting: bio-rigor (rigor guardrail) · seqeron-discovery (tool lookup) · bio-annotation (GC-skew/replication-origin) · seqeron-structural-variants (variant-level read-evidence SV/CNV).