@datagrok/peptides 1.27.9 → 1.27.10

package/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Peptides changelog
 
+## 1.27.10 (2026-07-08)
+
+### Features
+
+* Peptide Generation: New viewer that designs new candidate peptides from monomer-position statistics using an additive positional model, ranked by predicted activity (High/Low target) with per-peptide statistical support and novelty flags. Reuses the SAR viewers'/model's cached statistics when available, reacts to the dataframe filter, applies empirical-Bayes shrinkage to damp low-support noise, and carries an info icon explaining the model and its caveats. Appends per-position "basis" columns that render each chosen monomer (colored from the monomer library, width-fitted) with an effect circle (size = per-row-normalized effect magnitude, color = direction × significance: red raises, blue lowers, pale ≈ weak, grey = untestable) and the signed contribution value, with Most-Potent-Residues-style activity-distribution tooltips that explain the encoding. A header "+" button opens the generated peptides as a standalone table view, flattening the basis columns to "monomer, contribution, p-value" text. Not added to SAR analysis by default — add it manually from the viewers menu.
+
+* Moved the Peptides Playwright E2E suite into the package (playwright/); helpers from @datagrok-libraries/test/src/playwright
 ## 1.27.4 - 1.27.7 (2026-03-20)
 
 ### Features

package/CLAUDE.md

@@ -80,6 +80,7 @@ Key types:
 | `mutation-cliffs-viewer.ts` | `MutationCliffsViewer` | Line chart of mutation cliffs at a selected position. Splits by series column, syncs selection with main dataframe |
 | `cluster-max-activity-viewer.ts` | `ClusterMaxActivityViewer` | Scatter plot: cluster size vs max activity per cluster. Draws threshold lines, supports auto-selection of top quadrants |
 | `position-statistics-viewer.ts` | `SequencePositionStatsViewer` | Box/violin plot of numerical values grouped by monomers at a selected position. Configurable motif overhang |
+| `peptide-generation-viewer.ts` | `PeptideGenerationViewer` | **Optional, not added by default.** De novo peptide generator. Reuses `calculateMonomerPositionStatistics` (sharing the SAR viewers'/model's cached stats when they match) and an additive positional (Free-Wilson-style) model with empirical-Bayes shrinkage to synthesize new candidate peptides via bounded beam search, ranked by predicted activity (High/Low). Each row reports predicted activity, confidence, min support, mean p-value, significant positions and novelty. Appends per-position **basis columns** custom-rendered via `onCellRender` (library-colored width-fitted monomer + effect circle whose size = per-row-normalized effect magnitude and color = direction × significance (red raises / blue lowers, pale = weak, grey = untestable) + signed contribution number) with `showTooltip` activity-distribution tooltips that explain the encoding, mirroring Most Potent Residues. Builds sequences with `seqHandler.joiner` + `StringListSeqSplitted` + `getNewColumnFromList` |
 
 ### Export Features (context menu on Sequence Variability Map / Most Potent Residues)