Allow GenotypeGVCFHandler to leave a memory buffer for C layer

Author: bbimber

SequenceAnalysis/src/org/labkey/sequenceanalysis/analysis/GenotypeGVCFHandler.java

@@ -394,6 +394,20 @@ private File runGenotypeGVCFs(PipelineJob job, JobContext ctx, ProcessVariantsHa
                     toolParams.add("--genomicsdb-shared-posixfs-optimizations");
                 }
 
+                Integer maxRam = SequencePipelineService.get().getMaxRam();
+                int nativeMemoryBuffer = ctx.getParams().optInt("variantCalling.GenotypeGVCFs.nativeMemoryBuffer", 0);
+                if (maxRam != null && nativeMemoryBuffer > 0)
+                {
+                    ctx.getLogger().info("Adjusting RAM based on memory buffer (" + nativeMemoryBuffer + ")");
+                    maxRam = maxRam - nativeMemoryBuffer;
+
+                    if (maxRam < 1)
+                    {
+                        throw new PipelineJobException("After adjusting for nativeMemoryBuffer, maxRam is less than 1: " + maxRam);
+                    }
+                    wrapper.setMaxRamOverride(maxRam);
+                }
+
                 wrapper.execute(genome.getWorkingFastaFile(), outputVcf, toolParams, inputVcf);
                 try
                 {

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

@@ -198,31 +198,16 @@ public String getFileSuffix()
 
     private void findAdditionalData(SequenceOutputFile loupeFile, CSVWriter writer, PipelineJob job) throws IOException, PipelineJobException
     {
-        Set<Integer> hashingReadsets = new HashSet<>();
         Set<Integer> citeReadsets = new HashSet<>();
         Container targetContainer = job.getContainer().isWorkbook() ? job.getContainer().getParent() : job.getContainer();
         TableSelector ts = new TableSelector(QueryService.get().getUserSchema(job.getUser(), targetContainer, SingleCellSchema.NAME).getTable(SingleCellSchema.TABLE_CDNAS), PageFlowUtil.set("hashingReadsetId", "citeseqReadsetId"), new SimpleFilter(FieldKey.fromString("readsetId"), loupeFile.getReadset()), null);
         ts.forEachResults(rs -> {
-            if (rs.getObject(FieldKey.fromString("hashingReadsetId")) != null)
-            {
-                hashingReadsets.add(rs.getInt(FieldKey.fromString("hashingReadsetId")));
-            }
-
             if (rs.getObject(FieldKey.fromString("citeseqReadsetId")) != null)
             {
                 citeReadsets.add(rs.getInt(FieldKey.fromString("citeseqReadsetId")));
             }
         });
 
-        if (hashingReadsets.size() > 1)
-        {
-            throw new PipelineJobException("More than one hashing readset associated with GEX readset: " + loupeFile.getReadset());
-        }
-        else if (hashingReadsets.size() == 1)
-        {
-            writeExtraData(loupeFile.getRowid(), hashingReadsets.iterator().next(), job, "Cell Hashing Counts", writer, "HTO");
-        }
-
         if (citeReadsets.size() > 1)
         {
             throw new PipelineJobException("More than one CITE-seq readset associated with GEX readset: " + loupeFile.getReadset());

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

@@ -32,7 +32,6 @@ public Provider()
             ), null, null);
         }
 
-
         @Override
         public MergeSeurat create(PipelineContext ctx)
         {