@platforma-open/milaboratories.sort-seq-analysis.block 1.0.6 → 1.1.0

package/block-pack/CHANGELOG.md

@@ -1,5 +1,25 @@
 # @platforma-open/milaboratories.sort-seq-analysis.block
 
+## 1.1.0
+
+### Minor Changes
+
+- 6d40a83: Optional per-gate enrichment, nucleotide-level scores, renamed score columns
+
+  - **Per-gate enrichment**. Name the gate value holding your unsorted library and
+    the run emits one enrichment per condition and gate. Ordering the gates and naming an input are
+    independent, so a run can produce either set of scores or both. A variant missing from a gate
+    scores 0, and for the enrichment the read-count floor counts the variant's input reads.
+  - **Nucleotide datasets** are scored per nucleotide variant and rolled up per protein by pooling
+    their reads. Each protein enrichment carries an **error** in the same units: the larger of the
+    read-count error and the disagreement between the protein's nucleotide variants.
+  - The two score columns are renamed to the words the field uses: **Mean bin** and **Mean bin vs
+    parent**. Column names and values are unchanged.
+  - **Breaking for downstream picks:** both score columns now carry a `pl7.app/alphabet` domain
+    (`aminoacid` or `nucleotide`), so the protein and nucleotide levels of one run can be told
+    apart. A block that picked one of them before the upgrade, such as the Mutation Explorer's
+    score list, must pick it again after this block re-runs.
+
 ## 1.0.6
 
 ### Patch Changes

package/block-pack/description.md

@@ -1,18 +1,47 @@
 ## Overview
 
-A sort-seq experiment hands you a library of protein variants and, for each gate the sorter collected, a
-count of how many reads of each variant landed in it. Nothing in that table is a binding measurement: it is
-a spread of reads across gates, and the gates' order along the binding axis is the only thing that makes one
-variant's spread better than another's.
+A sort-seq experiment hands you a library of protein variants. For each gate the sorter collected, you get a
+count of how many reads of each variant landed in it. Nothing in that table is a binding measurement. It is a
+spread of reads across gates, and the block turns that spread into a score.
 
-This block turns that spread into a score, one per condition. For every variant it emits:
+Two things decide what a run produces, and they are independent:
 
-- **Gate rank mean** — the read-weighted mean of the gate ranks the variant's reads fell in, in gate-rank
-  units. Higher means the variant sorted into higher gates. Not an affinity and not calibrated.
-- **Bin score** — the gate rank mean minus the parent's. Zero means the variant behaves like the parent at
+- **Are the gates ordered?** An order along the binding axis is what makes one variant's spread better than
+  another's.
+- **Is one gate an unsorted input?** An unsorted sample is a reference each gate can be compared against.
+
+A run may have either, or both. With both, the block emits both sets of scores.
+
+## Ordered Gates: Mean Bin
+
+When the gates are ordered, every variant gets two values per condition:
+
+- **Mean bin** — the read-weighted mean of the gate ranks the variant's reads fell in, in gate-rank units.
+  Higher means the variant sorted into higher gates. Not an affinity and not calibrated.
+- **Mean bin vs parent** — the mean bin minus the parent's. Zero means the variant behaves like the parent at
   that condition.
 
-A run over N conditions emits both quantities N times, once per condition. A one-condition run is an
-ordinary run: it gets both quantities, with the run's single condition on each column exactly as a
-two-condition run would carry two.
+## An Unsorted Input: Per-Gate Enrichment
+
+Name the gate value holding your unsorted library, and every variant gets an **enrichment** for each gate:
+its share of reads in that gate divided by its share in the input. A variant the gate did not collect
+gets 0, and for the enrichment the read-count floor counts the variant's reads in the input.
+
+The enrichment is computed for every gate the run covers, ordered or not. So it works both on gates that
+only differ in what they select, and on gates that sit along a binding axis. An ordered run with an input
+gets an enrichment per gate on top of its mean-bin scores.
+
+## Nucleotide-Level Data
+
+When the dataset carries nucleotide variants, the block scores each one and then rolls them up per protein.
+A protein's score pools the reads of all its nucleotide variants. Beside each protein enrichment sits an
+**error** in the same units: the larger of the read-count error and the disagreement between the
+protein's nucleotide variants, together with how many variants it was measured on.
+
+The two levels are shown as two tables, because they sit on different axes and have different row counts.
+
+## Conditions
 
+A run over N conditions emits every quantity N times, once per condition. A one-condition run is an ordinary
+run. It gets the same columns, with the run's single condition on each, exactly as a two-condition run would
+carry two.

package/block-pack/manifest.json

@@ -1 +1 @@
-{"schema":"v2","description":{"id":{"organization":"milaboratories","name":"sort-seq-analysis","version":"1.0.6"},"components":{"workflow":{"type":"workflow-v1","main":{"type":"relative","path":"main.plj.gz"}},"model":{"type":"relative","path":"model.json"},"ui":{"type":"relative","path":"ui.tgz"}},"meta":{"title":"Sort-Seq Analysis","description":"Scores protein variants from a sort-seq (FACS bin) experiment: per condition, the read-weighted mean of the gate ranks each variant sorted into, and that value minus the parent's.","longDescription":{"type":"relative","path":"description.md"},"changelog":{"type":"relative","path":"CHANGELOG.md"},"logo":{"type":"relative","path":"block-logo.png"},"url":"https://github.com/platforma-open/sort-seq-analysis","support":"mailto:support@milaboratories.com","tags":["downstream","dms","antibody","assay"],"organization":{"name":"MiLaboratories Inc","url":"https://milaboratories.com/","logo":{"type":"relative","path":"organization-logo.png"}}},"featureFlags":{"supportsLazyState":true,"supportsPframeQueryRanking":true,"requiresUIAPIVersion":3,"requiresModelAPIVersion":2,"requiresCreatePTable":2,"requiresPFramesVersion":1001031,"requiresPFrameSpec":true,"requiresPFrame":true,"requiresDialog":true,"requiresColumnsCollection":true},"kind":"@platforma-open/milaboratories.sort-seq-analysis.kind@1.0.1"},"timestamp":1789119499571,"files":[{"name":"main.plj.gz","size":798776,"sha256":"1F72DB6258B50FC57D229D1AFD387A9FAE7DAE78FB2FE14214B84CE206CDEE55"},{"name":"model.json","size":558031,"sha256":"013E485766BB7764C99E786F7DA9D95265B01BD7EE4A0C57FA1B50C6F4C74EAB"},{"name":"ui.tgz","size":6371759,"sha256":"8EA11E0FC3886EE1C76B4558EC571655BD844AF37B38D1D332AEF274A2425705"},{"name":"organization-logo.png","size":24439,"sha256":"FA71390C77C91E4B7FAAE5640D00F92F1E3F2869296F68B6040DD7CC549A50B5"},{"name":"description.md","size":1035,"sha256":"49CEBEA9C9B395A12ECC5D1FF7CDEE9E6BA371C5D4E6E7FA874C07A3DC1AB3E1"},{"name":"CHANGELOG.md","size":4809,"sha256":"A344B58883D423E6B95DAE15C2E9C1B943151832B5F9D38E16896BB7A39E3BFB"},{"name":"block-logo.png","size":21527,"sha256":"6BB33BAF0CD039549661B51AE490373BE60D1811EC71F5023400928293BC2427"}]}
+{"schema":"v2","description":{"id":{"organization":"milaboratories","name":"sort-seq-analysis","version":"1.1.0"},"components":{"workflow":{"type":"workflow-v1","main":{"type":"relative","path":"main.plj.gz"}},"model":{"type":"relative","path":"model.json"},"ui":{"type":"relative","path":"ui.tgz"}},"meta":{"title":"Sort-Seq Analysis","description":"Scores protein variants from a sort-seq (FACS bin) experiment. Ordered gates give each variant a read-weighted mean gate rank and its difference from the parent's. An unsorted input gate adds a per-gate enrichment, on ordered or unordered gates.","longDescription":{"type":"relative","path":"description.md"},"changelog":{"type":"relative","path":"CHANGELOG.md"},"logo":{"type":"relative","path":"block-logo.png"},"url":"https://github.com/platforma-open/sort-seq-analysis","support":"mailto:support@milaboratories.com","tags":["downstream","dms","antibody","assay"],"organization":{"name":"MiLaboratories Inc","url":"https://milaboratories.com/","logo":{"type":"relative","path":"organization-logo.png"}}},"featureFlags":{"supportsLazyState":true,"supportsPframeQueryRanking":true,"requiresUIAPIVersion":3,"requiresModelAPIVersion":2,"requiresCreatePTable":2,"requiresPFramesVersion":1001031,"requiresPFrameSpec":true,"requiresPFrame":true,"requiresDialog":true,"requiresColumnsCollection":true},"kind":"@platforma-open/milaboratories.sort-seq-analysis.kind@1.0.1"},"timestamp":1790183853587,"files":[{"name":"main.plj.gz","size":811206,"sha256":"8F249AC886A648275F5D675C11BBEC5877C9E7D53EF9A4CC4D15E6EEFF78A546"},{"name":"model.json","size":567097,"sha256":"575503B818163120E48B98F57213E34B3FBAE9072B178CA824E198A369B07329"},{"name":"ui.tgz","size":6382446,"sha256":"049FA9A17FE89E4B6A6F0FDD89C110BD474DC4916B00403F70F9F265FDFB28D7"},{"name":"organization-logo.png","size":24439,"sha256":"FA71390C77C91E4B7FAAE5640D00F92F1E3F2869296F68B6040DD7CC549A50B5"},{"name":"description.md","size":2503,"sha256":"50CE6B002C3D3B3FBD450298F4FDE4E5C7B2FE4EA6ADAA932098632F2C696415"},{"name":"CHANGELOG.md","size":6082,"sha256":"6B8BC2380AF2E2CA0C29878C65970A051EA5C20EBAD772F512A38E8C2C3661CD"},{"name":"block-logo.png","size":21527,"sha256":"6BB33BAF0CD039549661B51AE490373BE60D1811EC71F5023400928293BC2427"}]}