Add support for running VISION

Author: bbimber

singlecell/resources/chunks/RunVision.R

@@ -0,0 +1,14 @@
+visionFiles <- data.frame(DatasetId = character(), FileName = character())
+
+for (datasetId in names(seuratObjects)) {
+    printName(datasetId)
+    seuratObj <- readRDS(seuratObjects[[datasetId]])
+
+    visionObj <- CellMembrane::RunVisionForMSigDB(seuratObj, metadataCols = metadataCols)
+    fileName <- paste0('vision.', datasetId, '.rds')
+    saveRDS(visionObj, file = fileName)
+
+    visionFiles <- rbind(visionFiles, data.frame(DatasetId = datasetId, FileName = fileName))
+}
+
+write.table(visionFiles, file = 'visionFiles.txt', row.names = FALSE, col.names = FALSE, quote = FALSE, sep = '\t')

singlecell/src/org/labkey/singlecell/SingleCellModule.java

@@ -184,6 +184,7 @@ public static void registerPipelineSteps()
         SequencePipelineService.get().registerPipelineStep(new TrainCelltypist.Provider());
         SequencePipelineService.get().registerPipelineStep(new CheckExpectations.Provider());
         SequencePipelineService.get().registerPipelineStep(new CommonFilters.Provider());
+        SequencePipelineService.get().registerPipelineStep(new RunVision.Provider());
 
         SequenceAnalysisService.get().registerFileHandler(new NimbleAlignmentStep());
     }

singlecell/src/org/labkey/singlecell/pipeline/singlecell/FindMarkers.java

@@ -45,6 +45,12 @@ public Provider()
                     SeuratToolParameter.create("foldChangeThreshold", "Log2 Fold-Change Threshold", "Only genes with log2-foldchange above this will be reported", "ldk-numberfield", new JSONObject(){{
                         put("minValue", 0);
                     }}, 0.25),
+                    SeuratToolParameter.create("minPct", "Min Pct", "Only test genes that are detected in a minimum fraction of min.pct cells in either of the two populations", "ldk-numberfield", new JSONObject(){{
+                        put("minValue", 0);
+                    }}, 0.1),
+                    SeuratToolParameter.create("minDiffPct", "Min Diff Pct", "Only test genes that show a minimum difference in the fraction of detection between the two groups", "ldk-numberfield", new JSONObject(){{
+                        put("minValue", 0);
+                    }}, null, "minDiffPct", false),
                     getSeuratThreadsParam()
                 ), Arrays.asList("/sequenceanalysis/field/TrimmingTextArea.js"), null);
         }

singlecell/src/org/labkey/singlecell/pipeline/singlecell/RunVision.java

@@ -0,0 +1,114 @@
+package org.labkey.singlecell.pipeline.singlecell;
+
+import au.com.bytecode.opencsv.CSVReader;
+import org.json.JSONObject;
+import org.labkey.api.pipeline.PipelineJobException;
+import org.labkey.api.reader.Readers;
+import org.labkey.api.sequenceanalysis.SequenceOutputFile;
+import org.labkey.api.sequenceanalysis.pipeline.AbstractPipelineStepProvider;
+import org.labkey.api.sequenceanalysis.pipeline.PipelineContext;
+import org.labkey.api.sequenceanalysis.pipeline.SequenceOutputHandler;
+import org.labkey.api.singlecell.pipeline.SeuratToolParameter;
+import org.labkey.api.singlecell.pipeline.SingleCellStep;
+
+import java.io.File;
+import java.io.IOException;
+import java.util.Arrays;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+public class RunVision extends AbstractCellMembraneStep
+{
+    public RunVision(PipelineContext ctx, RunVision.Provider provider)
+    {
+        super(provider, ctx);
+    }
+
+    public static class Provider extends AbstractPipelineStepProvider<SingleCellStep>
+    {
+        public Provider()
+        {
+            super("RunVision", "Run VISION", "VISION", "This will run VISION, a tool for functional interpretation of scRNA-seq data, saving the result to an RDS file.", Arrays.asList(
+                    SeuratToolParameter.create("metadataCols", "Field(s)", "This will subset the seuratObj to just these columns prior to running VISION.", "sequenceanalysis-trimmingtextarea", new JSONObject(){{
+                        put("allowBlank", false);
+                        put("height", 200);
+                        put("delimiter", ",");
+                    }}, "nCount_RNA,ClusterNames_0.2,ClusterNames_0.4,ClusterNames_0.6,ClusterNames_0.8", "metadataCols", true, true).delimiter(",")
+                ), Arrays.asList("/sequenceanalysis/field/TrimmingTextArea.js"), null);
+        }
+
+        @Override
+        public RunVision create(PipelineContext ctx)
+        {
+            return new RunVision(ctx, this);
+        }
+    }
+
+    @Override
+    public boolean createsSeuratObjects()
+    {
+        return false;
+    }
+
+    @Override
+    public String getFileSuffix()
+    {
+        return "vision";
+    }
+
+    @Override
+    public Output execute(SequenceOutputHandler.JobContext ctx, List<SeuratObjectWrapper> inputObjects, String outputPrefix) throws PipelineJobException
+    {
+        Output output = super.execute(ctx, inputObjects, outputPrefix);
+
+        File saved = new File(ctx.getOutputDir(), "visionFiles.txt");
+        if (!saved.exists())
+        {
+            throw new PipelineJobException("Unable to find file: " + saved.getPath());
+        }
+
+        try (CSVReader reader = new CSVReader(Readers.getReader(saved), '\t'))
+        {
+            String[] line;
+            while ((line = reader.readNext()) != null)
+            {
+                File rds = new File(ctx.getOutputDir(), line[1]);
+                if (!rds.exists())
+                {
+                    throw new PipelineJobException("Unable to find file: " + rds.getPath());
+                }
+
+                final String datasetId = line[0];
+                Set<SeuratObjectWrapper> wrappers = inputObjects.stream().filter(x -> datasetId.equals(x.getDatasetId())).collect(Collectors.toSet());
+                if (wrappers.size() == 0)
+                {
+                    throw new PipelineJobException("Unable to find seurat object wrapper for: " + datasetId);
+                }
+                else if (wrappers.size() > 1)
+                {
+                    throw new PipelineJobException("More than one seurat object wrapper matched: " + datasetId + ", found: " + wrappers.stream().map(SeuratObjectWrapper::getDatasetId).collect(Collectors.joining(", ")));
+                }
+
+                SeuratObjectWrapper wrapper = wrappers.iterator().next();
+
+                SequenceOutputFile so = new SequenceOutputFile();
+                so.setFile(rds);
+                so.setCategory("VISION Analysis");
+                so.setLibrary_id(ctx.getSequenceSupport().getCachedGenomes().iterator().next().getGenomeId());
+                so.setReadset(wrapper.getReadsetId());
+                so.setName(wrapper.getDatasetName() == null ? wrapper.getDatasetId() : wrapper.getDatasetName() + ": VISION Analysis");
+
+                ctx.getFileManager().addSequenceOutput(so);
+            }
+        }
+        catch (IOException e)
+        {
+            throw new PipelineJobException(e);
+        }
+
+        ctx.getFileManager().addIntermediateFile(saved);
+
+        return output;
+    }
+}