dsh-molbio-tools 0.22.2 → 0.24.1

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "dsh-molbio-tools",
-  "version": "0.22.2",
-  "description": "Zero-dependency molecular-biology plugin bundle for the DeepSeek Harness: sequence analysis, primer design (incl. cross-intron qPCR primers, TaqMan hydrolysis/MGB probes with ordering CSV and multiplex compatibility), in-silico PCR, cloning simulation, SnapGene .dna parsing and plasmid maps, Sanger verification, protein tools (incl. helical wheel and hydropathy plots), methylation-aware restriction analysis with double-digest buffer checks, qPCR analysis and plotting, multiple sequence alignment that conserves every input residue, conservation analysis, sequence logos, codon-usage analysis, phylogenetic trees with bootstrap support, Newick I/O and long-name label wrapping, CpG-island/GC-skew composition analysis, read-level FASTQ QC, batch file analysis with CSV export, silent restriction-site mutagenesis, family-aware CRISPR gRNA design (SpCas9/SaCas9/Cas12a/Cas13) with off-target search, an opt-in hand-off that shows a generated plot to the model as an attached image, plus a browser-side right-sidebar panel (plasmid maps and sequence logos drawn in the Web GUI) and a literature/protocol/experiment record system.",
+  "version": "0.24.1",
+  "description": "Zero-dependency molecular-biology plugin bundle for the DeepSeek Harness: sequence analysis, primer design (incl. cross-intron qPCR primers, TaqMan hydrolysis/MGB probes with ordering CSV and multiplex compatibility), in-silico PCR, cloning simulation, SnapGene .dna parsing and plasmid maps, Sanger verification, protein tools (incl. helical wheel, hydropathy and charge-pH plots), methylation-aware restriction analysis with double-digest buffer checks, qPCR analysis and plotting, multiple sequence alignment that conserves every input residue, conservation analysis, sequence logos, codon-usage analysis, phylogenetic trees with bootstrap support, Newick I/O and long-name label wrapping, CpG-island/GC-skew composition analysis, k-mer dotplots, read-level FASTQ QC, vector/adapter contamination screening, FASTQ trimming/filtering, batch file analysis with CSV export, silent restriction-site mutagenesis, family-aware CRISPR gRNA design (SpCas9/SaCas9/Cas12a/Cas13) with off-target search, an opt-in hand-off that shows a generated plot to the model as an attached image, plus a browser-side right-sidebar panel (plasmid maps and sequence logos drawn in the Web GUI) and a literature/protocol/experiment record system.",
   "type": "module",
   "main": "index.mjs",
   "engines": {
@@ -59,6 +59,10 @@
     "sanger.mjs",
     "protein.mjs",
     "protein-structure.mjs",
+    "protein-charge.mjs",
+    "dotplot.mjs",
+    "contamination.mjs",
+    "fastq-trim.mjs",
     "taqman.mjs",
     "multiplex.mjs",
     "methylation.mjs",
@@ -98,7 +102,8 @@
         "0.2.1-rc.2": "compatible",
         "0.2.1-rc.3": "compatible",
         "0.2.1-rc.4": "unknown",
-        "0.2.1-rc.5": "compatible"
+        "0.2.1-rc.5": "compatible",
+        "0.2.2-rc.1": "compatible"
       }
     },
     "bundle": {

package/papers.mjs

@@ -51,6 +51,12 @@ export async function loadLibrary(fs, path) {
     if (typeof parsed !== 'object' || parsed === null || !Array.isArray(parsed.papers)) {
       throw new Error('expected an object with a "papers" array');
     }
+    // A hand-edited library can contain a null or scalar entry. Without this
+    // check, paperKey()/render() would throw a raw TypeError; a clear input
+    // error names the file instead.
+    if (!parsed.papers.every((paper) => typeof paper === 'object' && paper !== null && !Array.isArray(paper))) {
+      throw new Error('every entry in "papers" must be an object');
+    }
     return parsed;
   } catch (error) {
     throw new MolbioInputError(`the paper library at ${path} is not valid JSON: ${error instanceof Error ? error.message : String(error)}`);
@@ -115,6 +121,10 @@ export function updatePaper(library, id, changes) {
   if (Array.isArray(changes.tags)) {
     paper.tags = changes.tags.filter((tag) => typeof tag === 'string').map((tag) => tag.trim()).filter((tag) => tag !== '');
   }
+  // The id is DERIVED from pmid, then url, then title+year. Recompute it after
+  // an update: leaving the old title-based id in place means adding the renamed
+  // paper later is not recognised as a duplicate.
+  paper.id = paperKey(paper);
   return paper;
 }
 

package/plot.mjs

@@ -70,6 +70,20 @@ function validateNumbers(values, label) {
   return values;
 }
 
+/** Smallest datum, by loop: spreading a long series into Math.min overflows the argument limit. */
+function minValue(values) {
+  let min = Infinity;
+  for (const value of values) if (value < min) min = value;
+  return min;
+}
+
+/** Largest datum, by loop (same reason as {@link minValue}). */
+function maxValue(values) {
+  let max = -Infinity;
+  for (const value of values) if (value > max) max = value;
+  return max;
+}
+
 function frameParts(title, xLabel, yLabel) {
   const parts = [];
   parts.push(`<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 ${WIDTH} ${HEIGHT}" font-family="${FIGURE_FONT}" role="img">`);
@@ -110,11 +124,13 @@ export function renderBarChart({ title, x_label, y_label, labels, values, errors
     }
     errs = errors;
   }
-  let min = Math.min(0, ...vals);
-  let max = Math.max(0, ...vals);
+  let min = Math.min(0, minValue(vals));
+  let max = Math.max(0, maxValue(vals));
   if (errs !== undefined) {
-    min = Math.min(min, ...vals.map((v, i) => v - errs[i]));
-    max = Math.max(max, ...vals.map((v, i) => v + errs[i]));
+    for (let index = 0; index < vals.length; index++) {
+      min = Math.min(min, vals[index] - errs[index]);
+      max = Math.max(max, vals[index] + errs[index]);
+    }
   }
   if (max === min) max = min + 1;
   const parts = frameParts(title, x_label, y_label);
@@ -158,15 +174,15 @@ export function renderScatterChart({ title, x_label, y_label, x, y, fit }) {
     }
     fitLine = fit;
   }
-  let xMin = Math.min(...xs);
-  let xMax = Math.max(...xs);
+  let xMin = minValue(xs);
+  let xMax = maxValue(xs);
   if (xMax === xMin) {
     const pad = Math.abs(xMax) * 0.1 + 1;
     xMin -= pad;
     xMax += pad;
   }
-  let yMin = Math.min(...ys);
-  let yMax = Math.max(...ys);
+  let yMin = minValue(ys);
+  let yMax = maxValue(ys);
   if (fitLine !== undefined) {
     yMin = Math.min(yMin, fitLine.intercept + fitLine.slope * xMin, fitLine.intercept + fitLine.slope * xMax);
     yMax = Math.max(yMax, fitLine.intercept + fitLine.slope * xMin, fitLine.intercept + fitLine.slope * xMax);

package/preset/install.mjs

@@ -55,6 +55,14 @@ function flag(name) {
 }
 
 const profile = flag('profile') === undefined ? 'web' : String(flag('profile'));
+// DSH profile names are directory names under <dshHome>/profiles. Reject
+// separators/control characters before the value can reach a child process:
+// on Windows this script runs through a shell, where an unquoted metacharacter
+// (for example "web & calc") would be a command injection.
+if (!/^[A-Za-z0-9][A-Za-z0-9._-]*$/.test(profile) || profile === '.' || profile === '..') {
+  console.error(`invalid --profile ${JSON.stringify(profile)}: start with a letter or digit, then letters, digits, ".", "_" or "-"`);
+  process.exit(2);
+}
 const checkOnly = flag('check') === true;
 const dshHomeFlag = flag('dsh-home');
 const dshHome = resolve(

package/preset/molbio-lab/agent.cordis.yml

@@ -17,7 +17,7 @@
 #
 # Edit THIS file, then run `node build/preset-patch.mjs`.
 #
-# NOTE: the mode is the ONLY place these 60 tools load. The same bundle patch's
+# NOTE: the mode is the ONLY place these 64 tools load. The same bundle patch's
 # other insert (`molbio-client` → `./host.mjs`) deliberately registers nothing —
 # registering the tool at the host level as well would put it in every session
 # and is not what this package is for.
@@ -48,18 +48,20 @@
 # list carries no comments of its own, so none appear below; provenance and the
 # upgrade rules live in this header.
 #
-# Baseline: the `standard` preset of dsh 0.2.1-rc.3 (harness root
+# Baseline: the `standard` preset of dsh 0.2.2-rc.1 (harness root
 # `%APPDATA%\npm\node_modules\@deepseek-ai\dsh`, file
 # `node_modules/@deepseek-ai/dsh-web-app/presets/standard.patch.yml`, whose
 # rows live under the `preset-standard` row's `config.plugins:`).
 #
-# 0.2.1-rc.3 ships the SAME row list as 0.2.1-rc.2 (verified by diffing the
-# file), so no row changed in this bump. The last upstream row change was
-# 0.2.1-rc.2: it added `time-context`, `tool-schedule` and `tool-plugin-market`;
-# dropped the two disabled optional subagent providers (`tool-subagent-codex`,
-# `tool-subagent-claude-code`) and `tool-ralph`; dropped the persona `suffix` and
-# both `backgroundMode: continuable` settings; and reworded the plan-mode
-# section. Absorb such a bump verbatim — see workflow.md §4.
+# 0.2.2-rc.1 is the FIRST row change since 0.2.1-rc.2 dropped the optional
+# subagent providers and `tool-ralph`. It folded `list_agents` into
+# `@deepseek-ai/dsh-tool-subagent-control` (the `./list-agents` subpath export is
+# gone, so the `tool-subagent-list-agents` row was dropped here), renamed the
+# `tool-subagent` model-selection settings to `modelSelection` +
+# `modelDiscovery`, and made `applyPatch: false` explicit on `tool-fs`. A row
+# that names a deleted package subpath still PASSES a directory-existence check:
+# `test/preset-health.mjs` now resolves the specifier the way the Loader does.
+# Absorb such a bump verbatim — see workflow.md §4.
 #
 # Keeping the text verbatim is not cosmetic: a preset mounts DSH's OWN
 # packages, so upstream drift here kills the whole preset at mount (the failure
@@ -100,6 +102,11 @@
   disabled: !!js process.platform !== 'win32'
 - id: tool-fs
   name: '@deepseek-ai/dsh-tool-fs'
+  config:
+    # Set false to drop the apply_patch tool beside edit. (0.2.2-rc.1 makes this
+    # explicit in the shipped standard preset; the package default is false, and
+    # the row is kept verbatim so the two stay comparable.)
+    applyPatch: false
 - id: tool-fs-search
   name: '@deepseek-ai/dsh-tool-fs-search'
   config:
@@ -162,14 +169,13 @@
   config:
     - id: tool-subagent-control
       name: '@deepseek-ai/dsh-tool-subagent-control'
-    - id: tool-subagent-list-agents
-      name: '@deepseek-ai/dsh-tool-subagent-control/list-agents'
     - id: tool-subagent
       name: '@deepseek-ai/dsh-tool-subagent'
       config:
         provider: spawn
         toolName: subagent
-        modelSelectionSettings: true
+        modelSelection: true
+        modelDiscovery: true
     - id: tool-subagent-fork
       name: '@deepseek-ai/dsh-tool-subagent'
       config:
@@ -201,7 +207,7 @@
   name: '@deepseek-ai/dsh-plugin-market/tools'
   disabled: true
 
-# dsh-molbio-tools: molecular-biology toolset (60 tools across sequence
+# dsh-molbio-tools: molecular-biology toolset (64 tools across sequence
 # analysis, primers (incl. TaqMan probes and multiplex compatibility),
 # in-silico PCR, cloning (incl. Golden Gate), enzyme catalog, methylation-aware
 # and double digests, plasmids, Sanger verification, proteins (incl. helical

package/preset/molbio-lab/preset.yml

@@ -1,2 +1,2 @@
 name: Molecular Biology Lab
-description: 分子生物学专属模式：标准编码能力 + 60 个 molbio_* 工具（序列分析、引物设计（含跨内含子 qPCR、盐/浓度旋钮与目标位置偏好）、TaqMan 水解探针（含 MGB 与订购 CSV）与多重 PCR 互扰/退火温度/浓度建议、克隆模拟与选酶（含 Golden Gate）、酶目录查询、甲基化敏感位点与双酶切 buffer 兼容检查、多序列比对（每个输入残基都会进比对结果）、保守性分析与比对后处理（共识/缺口修剪/同一性矩阵/逐列覆盖度，含 sequence logo 图）、CRISPR gRNA 设计（SpCas9/SaCas9/Cas12a/Cas13，家族特定 PAM/PFS/种子）与脱靶搜索、SnapGene/GenBank 解析与质粒图谱、Sanger 测序验证、蛋白性质/密码子优化/螺旋轮投影与疏水性图、qPCR 与绘图、虚拟凝胶、实验台分析（FASTQ 质控、密码子使用、系统发生树（长叶名自动折行）、in-silico PCR、CpG/GC 组成、批量文件分析/CSV）、静默酶切位点突变设计——生成的 SVG 图片自动用系统默认应用打开，文献/协议/实验记录）。
+description: 分子生物学专属模式：标准编码能力 + 64 个 molbio_* 工具（序列分析、引物设计（含跨内含子 qPCR、盐/浓度旋钮与目标位置偏好）、TaqMan 水解探针（含 MGB 与订购 CSV）与多重 PCR 互扰/退火温度/浓度建议、克隆模拟与选酶（含 Golden Gate）、酶目录查询、甲基化敏感位点与双酶切 buffer 兼容检查、多序列比对（每个输入残基都会进比对结果）、保守性分析与比对后处理（共识/缺口修剪/同一性矩阵/逐列覆盖度，含 sequence logo 图）、CRISPR gRNA 设计（SpCas9/SaCas9/Cas12a/Cas13，家族特定 PAM/PFS/种子）与脱靶搜索、SnapGene/GenBank 解析与质粒图谱、Sanger 测序验证、蛋白性质/电荷-pH 曲线/密码子优化/螺旋轮投影与疏水性图、qPCR 与绘图、虚拟凝胶、实验台分析（FASTQ 质控、密码子使用、系统发生树（长叶名自动折行）、in-silico PCR、CpG/GC 组成、序列 dotplot、载体/接头污染筛查、FASTQ 修剪/过滤、批量文件分析/CSV）、静默酶切位点突变设计——生成的 SVG 图片自动用系统默认应用打开，文献/协议/实验记录）。

package/protein-charge.mjs

@@ -0,0 +1,158 @@
+/**
+ * dsh-molbio-tools/protein-charge.mjs
+ *
+ * Net-charge-vs-pH titration for a protein, with the pI found on the same
+ * curve. Pure calculation plus an SVG renderer: no external data, no services.
+ *
+ * The pK tables live in `protein.mjs` (PK_SETS) so `molbio_protein_props` and
+ * this tool can never disagree about the default set. `emboss` is offered as an
+ * explicitly named alternative; every result carries the set it used, because a
+ * pI is only meaningful together with its pK values.
+ */
+import { MolbioInputError } from './lib.mjs';
+import { PK_SETS, countAminoAcids, isoelectricPoint, netCharge, normalizeProtein } from './protein.mjs';
+import { axis, linearScale, niceTicks, panel, rect, series, svgDocument, textRun } from './svgio.mjs';
+
+/** Accepted ranges; the schema checks types and the module checks values. */
+export const CHARGE_LIMITS = {
+  maxLength: 20_000,
+  minPh: 0,
+  maxPh: 14,
+  minSteps: 2,
+  maxSteps: 141,
+};
+
+const round = (value, digits = 4) => Math.round(value * 10 ** digits) / 10 ** digits;
+
+/**
+ * Compute the titration curve and the pI for one protein.
+ *
+ * @param {string} raw protein sequence (one-letter codes).
+ * @param {{pkaSet?: string, phMin?: number, phMax?: number, steps?: number}} [options]
+ * @returns {{length: number, pka_set: string, pka_label: string, ph_min: number,
+ *   ph_max: number, steps: number, pI: number, charge_at_7_4: number,
+ *   charge_min: number, charge_max: number, points: Array<{ph: number, charge: number}>}}
+ */
+export function chargeCurve(raw, options = {}) {
+  const seq = normalizeProtein(raw);
+  if (seq.length > CHARGE_LIMITS.maxLength) {
+    throw new MolbioInputError('the protein is ' + seq.length + ' aa; the titration limit is ' + CHARGE_LIMITS.maxLength);
+  }
+  const pkaSet = options.pkaSet ?? 'bjellqvist';
+  const pk = PK_SETS[pkaSet];
+  if (pk === undefined) {
+    throw new MolbioInputError('unknown pka_set ' + JSON.stringify(pkaSet) + '; expected one of ' + Object.keys(PK_SETS).join(', '));
+  }
+  const phMin = options.phMin ?? 0;
+  const phMax = options.phMax ?? 14;
+  const steps = options.steps ?? 29;
+  if (!(typeof phMin === 'number' && Number.isFinite(phMin) && phMin >= CHARGE_LIMITS.minPh && phMin <= CHARGE_LIMITS.maxPh)) {
+    throw new MolbioInputError('ph_min must be a number between ' + CHARGE_LIMITS.minPh + ' and ' + CHARGE_LIMITS.maxPh);
+  }
+  if (!(typeof phMax === 'number' && Number.isFinite(phMax) && phMax >= CHARGE_LIMITS.minPh && phMax <= CHARGE_LIMITS.maxPh)) {
+    throw new MolbioInputError('ph_max must be a number between ' + CHARGE_LIMITS.minPh + ' and ' + CHARGE_LIMITS.maxPh);
+  }
+  if (!(phMax > phMin)) throw new MolbioInputError('ph_max must be greater than ph_min');
+  if (!Number.isInteger(steps) || steps < CHARGE_LIMITS.minSteps || steps > CHARGE_LIMITS.maxSteps) {
+    throw new MolbioInputError('steps must be an integer between ' + CHARGE_LIMITS.minSteps + ' and ' + CHARGE_LIMITS.maxSteps);
+  }
+  const counts = countAminoAcids(seq);
+  const points = [];
+  for (let index = 0; index < steps; index++) {
+    const ph = phMin + ((phMax - phMin) * index) / (steps - 1);
+    points.push({ ph: round(ph, 3), charge: round(netCharge(seq, counts, ph, pkaSet), 4) });
+  }
+  const charges = points.map((point) => point.charge);
+  return {
+    length: seq.length,
+    pka_set: pkaSet,
+    pka_label: pk.label,
+    ph_min: phMin,
+    ph_max: phMax,
+    steps,
+    pI: isoelectricPoint(seq, counts, pkaSet),
+    charge_at_7_4: round(netCharge(seq, counts, 7.4, pkaSet), 4),
+    charge_min: Math.min(...charges),
+    charge_max: Math.max(...charges),
+    points,
+  };
+}
+
+/**
+ * Render the titration curve as an SVG document: net charge against pH, with
+ * the zero line, the pI marker and the charged ends labelled.
+ *
+ * @param {ReturnType<typeof chargeCurve>} curve
+ * @param {{title?: string}} [options]
+ * @returns {string} SVG text.
+ */
+export function renderChargeCurve(curve, options = {}) {
+  const width = 720;
+  const height = 420;
+  const frame = panel({
+    x: 16,
+    y: 16,
+    width: width - 32,
+    height: height - 32,
+    title: options.title ?? 'Net charge vs pH',
+    subtitle: curve.length + ' aa · ' + curve.pka_label + ' pK set · pI ' + curve.pI,
+  });
+  const plot = frame.plot;
+  const yMin = Math.min(0, curve.charge_min);
+  const yMax = Math.max(0, curve.charge_max);
+  const pad = Math.max(0.5, (yMax - yMin) * 0.08);
+  const domain = [yMin - pad, yMax + pad];
+  const x = linearScale([curve.ph_min, curve.ph_max], [plot.x, plot.x + plot.width]);
+  const y = linearScale(domain, [plot.y + plot.height, plot.y]);
+  const parts = [frame.markup];
+  // Zero-charge line: the sign of the charge is the point of the plot.
+  if (domain[0] < 0 && domain[1] > 0) {
+    const zero = y.of(0);
+    parts.push('<line x1="' + plot.x + '" y1="' + zero + '" x2="' + (plot.x + plot.width) + '" y2="' + zero + '" stroke="#8c959f" stroke-width="1" stroke-dasharray="4 3"/>');
+  }
+  parts.push(axis({
+    from: plot.x,
+    to: plot.x + plot.width,
+    at: plot.y + plot.height,
+    ticks: niceTicks([curve.ph_min, curve.ph_max], 7).map((tick) => x.of(tick)),
+    orientation: 'x',
+    label: 'pH',
+    tickLabel: (pixel) => String(Math.round(x.invert(pixel) * 10) / 10),
+    title: 'pH',
+  }));
+  parts.push(axis({
+    from: plot.y,
+    to: plot.y + plot.height,
+    at: plot.x,
+    ticks: niceTicks(domain, 6).map((tick) => y.of(tick)),
+    orientation: 'y',
+    label: 'net charge',
+    tickLabel: (pixel) => String(Math.round(y.invert(pixel) * 10) / 10),
+    title: 'net charge (e)',
+  }));
+  parts.push(series({
+    values: curve.points.map((point) => point.charge),
+    xOf: (index) => x.of(curve.points[index].ph),
+    yOf: (charge) => y.of(charge),
+    stroke: '#0969da',
+    width: 2,
+  }));
+  if (curve.pI >= curve.ph_min && curve.pI <= curve.ph_max) {
+    const marker = x.of(curve.pI);
+    parts.push('<line x1="' + marker + '" y1="' + plot.y + '" x2="' + marker + '" y2="' + (plot.y + plot.height) + '" stroke="#bf3989" stroke-width="1" stroke-dasharray="2 3"/>');
+    parts.push(textRun({ x: marker + 4, y: plot.y + 12, text: 'pI ' + curve.pI, size: 9, fill: '#bf3989' }));
+  }
+  // Label the two ends of the curve with the charges they represent.
+  const first = curve.points[0];
+  const last = curve.points[curve.points.length - 1];
+  parts.push(textRun({ x: x.of(first.ph) + 4, y: y.of(first.charge) - 6, text: 'pH ' + first.ph + ': ' + first.charge, size: 9, fill: '#57606a' }));
+  parts.push(textRun({ x: x.of(last.ph) - 4, y: y.of(last.charge) + 14, text: 'pH ' + last.ph + ': ' + last.charge, size: 9, fill: '#57606a', anchor: 'end' }));
+  parts.push(rect({ x: plot.x, y: plot.y, width: plot.width, height: plot.height, fill: 'none', stroke: '#d0d7de' }));
+  return svgDocument({
+    width,
+    height,
+    title: options.title ?? 'Net charge vs pH',
+    description: 'Titration curve for a ' + curve.length + ' aa protein using the ' + curve.pka_label + ' pK set; pI ' + curve.pI + '.',
+    body: parts.join('\n'),
+  });
+}

package/protein-structure.mjs

@@ -278,14 +278,22 @@ export function hydropathyProfile(rawSequence, options = {}) {
   }
   if (current !== null) peaks.push(current);
   const total = points.reduce((sum, point) => sum + point.hydropathy, 0);
+  // Loop rather than Math.min(...points): the protein length is not capped and
+  // a long input would overflow the argument limit with a RangeError.
+  let minimumHydropathy = Infinity;
+  let maximumHydropathy = -Infinity;
+  for (const point of points) {
+    if (point.hydropathy < minimumHydropathy) minimumHydropathy = point.hydropathy;
+    if (point.hydropathy > maximumHydropathy) maximumHydropathy = point.hydropathy;
+  }
   return {
     length: sequence.length,
     window,
     threshold,
     gravy: sequence.length === 0 ? 0 : round2([...sequence].reduce((sum, aa) => sum + KYTE_DOOLITTLE[aa], 0) / sequence.length),
     mean_profile: round3(total / points.length),
-    minimum_hydropathy: round3(Math.min(...points.map((point) => point.hydropathy))),
-    maximum_hydropathy: round3(Math.max(...points.map((point) => point.hydropathy))),
+    minimum_hydropathy: round3(minimumHydropathy),
+    maximum_hydropathy: round3(maximumHydropathy),
     peaks: peaks.map((peak) => ({ ...peak, maximum: round3(peak.maximum), length: peak.end - peak.start + 1 })),
     points,
   };

package/protein.mjs

@@ -32,48 +32,70 @@ const RESIDUE_MASS = {
 };
 const WATER = 18.015;
 
-// Bjellqvist 1993 pK values (EMBOSS-style)
-const PK = {
-  Nterm: 8.6, Cterm: 3.6, C: 9.0, D: 4.05, E: 4.45, H: 6.04,
-  K: 10.8, R: 12.0, Y: 10.3,
+/**
+ * Selectable pK sets for the Henderson-Hasselbalch charge model. `bjellqvist`
+ * is the set `proteinProperties` has always used for its pI; `emboss` is the
+ * Epk.dat table EMBOSS ships with `iep`. They are NOT interchangeable — the
+ * reported set name travels with every charge/pI value.
+ */
+export const PK_SETS = {
+  bjellqvist: {
+    label: 'Bjellqvist 1993',
+    Nterm: 8.6, Cterm: 3.6, C: 9.0, D: 4.05, E: 4.45, H: 6.04,
+    K: 10.8, R: 12.0, Y: 10.3,
+  },
+  emboss: {
+    label: 'EMBOSS Epk.dat',
+    Nterm: 8.6, Cterm: 3.6, C: 8.5, D: 3.9, E: 4.1, H: 6.5,
+    K: 10.8, R: 12.5, Y: 10.1,
+  },
 };
 
+const PK = PK_SETS.bjellqvist;
+
 const HYDROPATHY = {
   A: 1.8, R: -4.5, N: -3.5, D: -3.5, C: 2.5, Q: -3.5, E: -3.5,
   G: -0.4, H: -3.2, I: 4.5, L: 3.8, K: -3.9, M: 1.9, F: 2.8,
   P: -1.6, S: -0.8, T: -0.7, W: -0.9, Y: -1.3, V: 4.2,
 };
 
-function countAminoAcids(seq) {
+export function countAminoAcids(seq) {
   const counts = {};
   for (const aa of seq) counts[aa] = (counts[aa] ?? 0) + 1;
   return counts;
 }
 
-function chargeAt(seq, counts, pH) {
+/**
+ * Net charge at a given pH by Henderson-Hasselbalch summation. The terminal
+ * groups count once each; K/R/H are positive, D/E/C/Y negative. `pkaSet` is a
+ * key of PK_SETS (or a set object), so callers can name the table they used.
+ */
+export function netCharge(seq, counts, pH, pkaSet = 'bjellqvist') {
+  const pk = typeof pkaSet === 'string' ? PK_SETS[pkaSet] : pkaSet;
+  if (pk === undefined) throw new MolbioInputError('unknown pK set ' + JSON.stringify(pkaSet) + '; expected one of ' + Object.keys(PK_SETS).join(', '));
   let charge = 0;
   const pos = (pKa) => 1 / (1 + Math.pow(10, pH - pKa));
   const neg = (pKa) => 1 / (1 + Math.pow(10, pKa - pH));
-  charge += pos(PK.Nterm);
-  for (const [aa, pKa] of [['K', PK.K], ['R', PK.R], ['H', PK.H]]) {
+  charge += pos(pk.Nterm);
+  for (const [aa, pKa] of [['K', pk.K], ['R', pk.R], ['H', pk.H]]) {
     if (counts[aa] !== undefined) charge += counts[aa] * pos(pKa);
   }
-  charge -= neg(PK.Cterm);
-  for (const [aa, pKa] of [['D', PK.D], ['E', PK.E], ['C', PK.C], ['Y', PK.Y]]) {
+  charge -= neg(pk.Cterm);
+  for (const [aa, pKa] of [['D', pk.D], ['E', pk.E], ['C', pk.C], ['Y', pk.Y]]) {
     if (counts[aa] !== undefined) charge -= counts[aa] * neg(pKa);
   }
   return charge;
 }
 
 /** Isoelectric point by bisection over the net-charge function. */
-export function isoelectricPoint(seq, counts) {
+export function isoelectricPoint(seq, counts, pkaSet = 'bjellqvist') {
   let lo = 0;
   let hi = 14;
-  let chargeLo = chargeAt(seq, counts, lo);
-  let chargeHi = chargeAt(seq, counts, hi);
+  let chargeLo = netCharge(seq, counts, lo, pkaSet);
+  let chargeHi = netCharge(seq, counts, hi, pkaSet);
   for (let i = 0; i < 60; i++) {
     const mid = (lo + hi) / 2;
-    const chargeMid = chargeAt(seq, counts, mid);
+    const chargeMid = netCharge(seq, counts, mid, pkaSet);
     if (Math.abs(chargeMid) < 1e-4) return Math.round(mid * 100) / 100;
     if (Math.sign(chargeMid) === Math.sign(chargeLo)) {
       lo = mid;

package/records.mjs

@@ -27,6 +27,11 @@ export async function loadRecords(fs, path, key) {
     if (typeof parsed !== 'object' || parsed === null || !Array.isArray(parsed[key])) {
       throw new Error(`expected an object with a "${key}" array`);
     }
+    // See papers.mjs: a null/scalar entry would otherwise reach
+    // record.name.trim() (protocols) as a raw TypeError.
+    if (!parsed[key].every((record) => typeof record === 'object' && record !== null && !Array.isArray(record))) {
+      throw new Error(`every entry in "${key}" must be an object`);
+    }
     return parsed;
   } catch (error) {
     throw new MolbioInputError(`${path} is not a valid record file: ${error instanceof Error ? error.message : String(error)}`);
@@ -54,7 +59,7 @@ export function addProtocol(records, entry) {
   if (typeof entry !== 'object' || entry === null || typeof entry.name !== 'string' || entry.name.trim() === '') {
     throw new MolbioInputError('each protocol needs a non-empty "name"');
   }
-  if (records.some((record) => record.name.trim().toLowerCase() === entry.name.trim().toLowerCase())) {
+  if (records.some((record) => typeof record?.name === 'string' && record.name.trim().toLowerCase() === entry.name.trim().toLowerCase())) {
     throw new MolbioInputError(`a protocol named ${JSON.stringify(entry.name)} already exists`);
   }
   const protocol = {
@@ -78,7 +83,17 @@ export function addProtocol(records, entry) {
 export function updateProtocol(records, id, changes) {
   const protocol = records.find((record) => record.id === id);
   if (protocol === undefined) return undefined;
-  if (typeof changes.name === 'string' && changes.name.trim() !== '') protocol.name = changes.name.trim();
+  if (typeof changes.name === 'string' && changes.name.trim() !== '') {
+    const nextName = changes.name.trim();
+    // addProtocol enforces unique names; update must too, or a rename can
+    // create a duplicate the next add refuses to explain.
+    if (records.some((record) => record !== protocol
+      && typeof record?.name === 'string'
+      && record.name.trim().toLowerCase() === nextName.toLowerCase())) {
+      throw new MolbioInputError(`a protocol named ${JSON.stringify(nextName)} already exists`);
+    }
+    protocol.name = nextName;
+  }
   if (typeof changes.category === 'string' && changes.category.trim() !== '') protocol.category = changes.category.trim();
   if (Array.isArray(changes.steps)) {
     protocol.steps = changes.steps.filter((step) => typeof step === 'string').map((step) => step.trim()).filter((step) => step !== '');

package/seqio.mjs

@@ -84,12 +84,35 @@ export function toFasta(entries) {
   return entries.map((entry) => `>${entry.id}${entry.description !== '' ? ' ' + entry.description : ''}\n${entry.sequence}`).join('\n') + '\n';
 }
 
+/** Render entries as four-line FASTQ text (quality is preserved verbatim). */
+export function toFastq(entries) {
+  return entries.map((entry) => {
+    if (typeof entry.quality !== 'string' || entry.quality.length !== entry.sequence.length) {
+      throw new MolbioInputError(`entry ${entry.id}: sequence and quality lengths differ`);
+    }
+    return `@${entry.id}${entry.description !== undefined && entry.description !== '' ? ' ' + entry.description : ''}\n${entry.sequence}\n+\n${entry.quality}`;
+  }).join('\n') + '\n';
+}
+
 /** Summary statistics over parsed entries. */
 export function entryStats(entries, qualityMode = false) {
-  const lengths = entries.map((entry) => entry.sequence.length);
+  // Count in a loop instead of spreading into Math.min/Math.max: a large FASTQ
+  // can hold hundreds of thousands of reads (and millions of quality
+  // characters), and spreading such an array overflows the engine's argument
+  // limit with a RangeError. The numbers are identical; only the failure mode
+  // changes.
+  let totalBases = 0;
+  let minLength = entries.length === 0 ? 0 : Infinity;
+  let maxLength = 0;
+  let longest = 0;
   let gc = 0;
   let at = 0;
   for (const entry of entries) {
+    const length = entry.sequence.length;
+    totalBases += length;
+    if (length < minLength) minLength = length;
+    if (length > maxLength) maxLength = length;
+    if (length > longest) longest = length;
     for (const base of entry.sequence) {
       if (base === 'G' || base === 'C') gc++;
       else if (base === 'A' || base === 'T') at++;
@@ -97,22 +120,34 @@ export function entryStats(entries, qualityMode = false) {
   }
   const stats = {
     entries: entries.length,
-    total_bases: lengths.reduce((sum, len) => sum + len, 0),
-    min_length: Math.min(...lengths),
-    max_length: Math.max(...lengths),
-    mean_length: Math.round((lengths.reduce((sum, len) => sum + len, 0) / lengths.length) * 100) / 100,
+    total_bases: totalBases,
+    min_length: minLength,
+    max_length: maxLength,
+    mean_length: entries.length === 0 ? 0 : Math.round((totalBases / entries.length) * 100) / 100,
     gc_percent: gc + at === 0 ? 0 : Math.round((gc / (gc + at)) * 1000) / 10,
   };
   if (qualityMode) {
-    const allQ = entries.flatMap((entry) => [...entry.quality].map((ch) => ch.charCodeAt(0) - 33));
-    if (allQ.length > 0) {
-      const mean = allQ.reduce((sum, q) => sum + q, 0) / allQ.length;
-      const low = allQ.filter((q) => q < 20).length / allQ.length;
-      stats.quality_mean = Math.round(mean * 100) / 100;
-      stats.quality_min = Math.min(...allQ);
-      stats.quality_max = Math.max(...allQ);
-      stats.low_quality_fraction = Math.round(low * 1000) / 1000;
-      const maxLen = Math.min(150, Math.max(...entries.map((e) => e.sequence.length)));
+    let qualitySum = 0;
+    let qualityCount = 0;
+    let lowCount = 0;
+    let qualityMin = Infinity;
+    let qualityMax = -Infinity;
+    for (const entry of entries) {
+      for (const ch of entry.quality) {
+        const q = ch.charCodeAt(0) - 33;
+        qualitySum += q;
+        qualityCount++;
+        if (q < 20) lowCount++;
+        if (q < qualityMin) qualityMin = q;
+        if (q > qualityMax) qualityMax = q;
+      }
+    }
+    if (qualityCount > 0) {
+      stats.quality_mean = Math.round((qualitySum / qualityCount) * 100) / 100;
+      stats.quality_min = qualityMin;
+      stats.quality_max = qualityMax;
+      stats.low_quality_fraction = Math.round((lowCount / qualityCount) * 1000) / 1000;
+      const maxLen = Math.min(150, longest);
       const perPosition = [];
       for (let pos = 0; pos < maxLen; pos++) {
         const values = entries.filter((e) => e.sequence.length > pos).map((e) => e.quality.charCodeAt(pos) - 33);

package/test/charts.mjs

@@ -284,6 +284,17 @@ test('histogram binning honours an explicit count and degenerates safely', () =>
   assert.ok(tiny.bin_count >= 1 && Number.isFinite(tiny.bin_width), 'a two-value sample still gets a usable bin');
 });
 
+test('a table at the row cap does not overflow the argument limit', () => {
+  // 130k exceeds the engine's argument limit for Math.min(...values); the
+  // chart helpers must use a loop. 130000 is below MAX_ROWS (200000), so the
+  // input is one the tool advertises as supported.
+  const values = Array.from({ length: 130000 }, (_, index) => index % 1000);
+  const shape = histogramBins(values, 10);
+  assert.equal(shape.bins.reduce((sum, bin) => sum + bin.count, 0), values.length);
+  assert.equal(shape.bins[0].start, 0);
+  assert.equal(shape.bins.at(-1).end, 999);
+});
+
 // ── CSV / TSV reader ────────────────────────────────────────────────────────
 
 test('the reader detects the delimiter and keeps quoted fields intact', () => {