Merge pull request #286 from BimberLab/24.3_fb_merge

Merge discvr-23.11 to discvr-24.3

Author: bbimber

SequenceAnalysis/src/org/labkey/sequenceanalysis/run/alignment/ParagraphStep.java

@@ -43,6 +43,12 @@ public ParagraphStep()
         ));
     }
 
+    @Override
+    public boolean doSplitJobs()
+    {
+        return true;
+    }
+
     @Override
     public boolean canProcess(SequenceOutputFile o)
     {

jbrowse/api-src/org/labkey/api/jbrowse/JBrowseService.java

@@ -56,4 +56,6 @@ public interface LuceneIndexDetector
 
         boolean isAvailable(Container c);
     }
+
+    abstract public void clearLuceneCacheEntry(File luceneIndexDir);
 }

jbrowse/src/org/labkey/jbrowse/JBrowseLuceneSearch.java

@@ -30,6 +30,7 @@
 import org.labkey.api.sequenceanalysis.SequenceAnalysisService;
 import org.labkey.api.sequenceanalysis.pipeline.SequencePipelineService;
 import org.labkey.api.settings.AdminConsole;
+import org.labkey.api.util.ContextListener;
 import org.labkey.api.util.PageFlowUtil;
 import org.labkey.api.util.SystemMaintenance;
 import org.labkey.api.view.WebPartFactory;
@@ -110,6 +111,8 @@ public void doStartupAfterSpringConfig(ModuleContext moduleContext)
 
         JBrowseService.get().registerFieldCustomizer(new JBrowseLuceneSearch.DefaultJBrowseFieldCustomizer());
         JBrowseService.get().registerGroupsProvider(new JBrowseLuceneSearch.TestJBrowseGroupProvider());
+
+        ContextListener.addShutdownListener(new JBrowseLuceneSearch.ShutdownHandler());
     }
 
     public static void registerPipelineSteps()

jbrowse/src/org/labkey/jbrowse/JBrowseModule.java

@@ -319,6 +319,12 @@ public void cacheDefaultQuery(User u, String sessionId, String trackId)
         luceneSearch.cacheDefaultQuery();
     }
 
+    @Override
+    public void clearLuceneCacheEntry(File luceneIndexDir)
+    {
+        JBrowseLuceneSearch.clearCacheForFile(luceneIndexDir);
+    }
+
     public static final class DefaultLuceneIndexDetector implements LuceneIndexDetector
     {
         @Override

jbrowse/src/org/labkey/jbrowse/JBrowseServiceImpl.java

@@ -70,17 +70,16 @@ file.create(trackerFile)
 print(paste0('Total lines in ', trackerFile, ' on job start:', length(readLines(trackerFile))))
 
 saveData <- function(seuratObj, datasetId) {
-    message(paste0('Saving dataset: ', datasetId, ' with ', ncol(seuratObj), ' cells'))
+    logger::log_info(paste0('Saving dataset: ', datasetId, ' with ', ncol(seuratObj), ' cells'))
     print(paste0('Saving dataset: ', datasetId))
     print(seuratObj)
 
     seuratObj <- .TestSplitLayers(seuratObj)
 
     datasetIdForFile <- makeLegalFileName(datasetId)
     fn <- paste0(outputPrefix, '.', datasetIdForFile, '.seurat.rds')
-    message(paste0('Filename: ', fn))
 
-    message(paste0('Saving RDS file: ', fn, ' with ', ncol(seuratObj), ' cells'))
+    logger::log_info(paste0('Saving RDS file: ', fn, ' with ', ncol(seuratObj), ' cells'))
     barcodeFile <- paste0(outputPrefix, '.', datasetIdForFile, '.cellBarcodes.csv')
     metaFile <- paste0(outputPrefix, '.', datasetIdForFile, '.seurat.meta.txt')
 

singlecell/resources/chunks/Functions.R

@@ -15,6 +15,11 @@ mergeBatch <- function(dat) {
         } else {
             toMerge[[datasetId]] <- readSeuratRDS(dat[[datasetId]])
         }
+
+        if (ncol(toMerge[[datasetId]]) == 1) {
+            logger::log_info(paste0('Dataset has single cell, skipping: ', datasetId))
+            toMerge[[datasetId]] <- NULL
+        }
     }
 
     if (!is.null(assaysToDrop)) {
@@ -28,6 +33,10 @@ mergeBatch <- function(dat) {
         }
     }
 
+    if (length(toMerge) == 0) {
+        stop('There were no passing seurat objects!')
+    }
+
     seuratObj <- CellMembrane::MergeSeuratObjs(toMerge, projectName = projectName, doGC = doDiet, errorOnBarcodeSuffix = errorOnBarcodeSuffix)
     return(seuratObj)
 }
@@ -61,6 +70,7 @@ if (length(seuratObjects) == 1) {
         unlink(mergedObjectFiles[[1]])
     } else {
         logger::log_info('performing final merge')
+        # TODO: check for single cell in object
         seuratObj <- readRDS(mergedObjectFiles[[1]])
         unlink(mergedObjectFiles[[1]])
 

singlecell/resources/chunks/MergeSeurat.R

@@ -27,6 +27,7 @@ for (datasetId in names(seuratObjects)) {
     })
 
     if (sum(matchingCells) == 0) {
+      print(paste0('No matching cells, skipping: ', datasetId))
       next
     }
 
@@ -40,6 +41,8 @@ for (datasetId in names(seuratObjects)) {
 
   if (all(is.null(cellsToKeep))) {
     print('There were no matching cells')
+  } else if (length(cellsToKeep) <= thresholdToKeep) {
+    print(paste0('Too few cells, skipping: ', datasetId))
   } else {
     print(paste0('Total passing cells: ', length(cellsToKeep)))
     seuratObj <- subset(seuratObj, cells = cellsToKeep)

singlecell/resources/chunks/TcrFilter.R

@@ -1063,7 +1063,7 @@ else if ("NotUsed".equals(val))
                 pf.setMaximumFractionDigits(2);
 
                 NumberFormat decimal = DecimalFormat.getNumberInstance();
-                decimal.setGroupingUsed(false);
+                decimal.setGroupingUsed(true);
 
                 descriptions.add("Total Cells: " + decimal.format(totalCells));
                 if (hashingUsed)

singlecell/src/org/labkey/singlecell/analysis/AbstractSingleCellHandler.java

@@ -21,12 +21,15 @@ public static class Provider extends AbstractPipelineStepProvider<SingleCellStep
     {
         public Provider()
         {
-            super("TcrFilter", "TCR-Based Filter", "CellMembrane/Seurat", "This will filter a seurat object based on TCR data.", Collections.singletonList(
+            super("TcrFilter", "TCR-Based Filter", "CellMembrane/Seurat", "This will filter a seurat object based on TCR data.", Arrays.asList(
                     SeuratToolParameter.create("cdr3s", "CDR3s To Keep", "A comma- or newline-delimited list of CDR3 sequences where locus prefixes the AA sequence (i.e. TRB:XXXXXX or TRA:YYYYYYY). Any cell matching any of these CDR3s will be kept. If that cell has multiple chains for a locus (i.e. 'CASSXXXXX,CASSYYYYY'), then only one of these needs to match for that cell to be kept. Also, all the input CDR3s should be single-chain (i.e. 'TRA:XXXXX', not 'TRA:XXXX,YYYY').", "sequenceanalysis-trimmingtextarea", new JSONObject(){{
                         put("height", 150);
                         put("delimiter", ",");
                         put("stripCharsRe", "/(^['\"]+)|(['\"]+$)/g");
-                    }}, null).delimiter(",")
+                    }}, null).delimiter(","),
+                    SeuratToolParameter.create("thresholdToKeep", "Min Cells To Keep", "If fewer than this many cells remain, the object will be discarded. This is primary present because seurat cannot easily accommodate objects with 1 cells", "ldk-integerfield", new JSONObject(){{
+                        put("minValue", 0);
+                    }}, 1)
             ), List.of("/sequenceanalysis/field/TrimmingTextArea.js"), null);
         }