PD-2863: Run one star instance

pipelines/skylab/multiome/Multiome.wdl

@@ -160,10 +160,10 @@ workflow Multiome {
         File gene_metrics_gex = Optimus.gene_metrics
         File? cell_calls_gex = Optimus.cell_calls
         File h5ad_output_file_gex = JoinBarcodes.gex_h5ad_file
-        Array[File?] multimappers_EM_matrix = Optimus.multimappers_EM_matrix
-        Array[File?] multimappers_Uniform_matrix = Optimus.multimappers_Uniform_matrix
-        Array[File?] multimappers_Rescue_matrix = Optimus.multimappers_Rescue_matrix
-        Array[File?] multimappers_PropUnique_matrix = Optimus.multimappers_PropUnique_matrix
+        File? multimappers_EM_matrix = Optimus.multimappers_EM_matrix
+        File? multimappers_Uniform_matrix = Optimus.multimappers_Uniform_matrix
+        File? multimappers_Rescue_matrix = Optimus.multimappers_Rescue_matrix
+        File? multimappers_PropUnique_matrix = Optimus.multimappers_PropUnique_matrix
         File? gex_aligner_metrics = Optimus.aligner_metrics
         File? library_metrics = Optimus.library_metrics
         File? mtx_files = Optimus.mtx_files

pipelines/skylab/optimus/Optimus.wdl

@@ -166,45 +166,25 @@ workflow Optimus {
       ubuntu_docker_path = ubuntu_docker_prefix + ubuntu_docker
   }
 
-  call FastqProcessing.FastqProcessing as SplitFastq {
-    input:
-      i1_fastq = i1_fastq,
-      r1_fastq = r1_fastq,
-      r2_fastq = r2_fastq,
-      whitelist = whitelist,
-      chemistry = tenx_chemistry_version,
-      sample_id = input_id,
-      read_struct = read_struct,
-      warp_tools_docker_path = docker_prefix + warp_tools_docker
-  }
-
-  scatter(idx in range(length(SplitFastq.fastq_R1_output_array))) {
-    call StarAlign.STARsoloFastq as STARsoloFastq {
+  call StarAlign.STARsoloFastq as STARsoloFastq {
       input:
-        r1_fastq = [SplitFastq.fastq_R1_output_array[idx]],
-        r2_fastq = [SplitFastq.fastq_R2_output_array[idx]],
+        r1_fastq = r1_fastq,
+        r2_fastq = r2_fastq,
         star_strand_mode = star_strand_mode,
         white_list = whitelist,
         tar_star_reference = tar_star_reference,
         chemistry = tenx_chemistry_version,
         counting_mode = counting_mode,
         count_exons = count_exons,
-        output_bam_basename = output_bam_basename + "_" + idx,
+        output_bam_basename = output_bam_basename,
         soloMultiMappers = soloMultiMappers,
         samtools_star_docker_path = docker_prefix + samtools_star,
         is_slidetags = is_slidetags
     }
-  }
-  call Merge.MergeSortBamFiles as MergeBam {
-    input:
-      bam_inputs = STARsoloFastq.bam_output,
-      output_bam_filename = output_bam_basename + ".bam",
-      sort_order = "coordinate",
-      picard_cloud_docker_path = docker_prefix + picard_cloud_docker
-  }
+
   call Metrics.CalculateGeneMetrics as GeneMetrics {
     input:
-      bam_input = MergeBam.output_bam,
+      bam_input = STARsoloFastq.bam_output,
       mt_genes = mt_genes,
       original_gtf = annotations_gtf,
       input_id = input_id,
@@ -213,7 +193,7 @@ workflow Optimus {
 
   call Metrics.CalculateCellMetrics as CellMetrics {
     input:
-      bam_input = MergeBam.output_bam,
+      bam_input = STARsoloFastq.bam_output,
       mt_genes = mt_genes,
       original_gtf = annotations_gtf,
       input_id = input_id,
@@ -222,13 +202,13 @@ workflow Optimus {
 
   call StarAlign.MergeStarOutput as MergeStarOutputs {
     input:
-      barcodes = STARsoloFastq.barcodes,
-      features = STARsoloFastq.features,
-      matrix = STARsoloFastq.matrix,
-      cell_reads = STARsoloFastq.cell_reads,
-      summary = STARsoloFastq.summary,
-      align_features = STARsoloFastq.align_features,
-      umipercell = STARsoloFastq.umipercell,
+      barcodes = [STARsoloFastq.barcodes],
+      features = [STARsoloFastq.features],
+      matrix = [STARsoloFastq.matrix],
+      cell_reads = [STARsoloFastq.cell_reads],
+      summary = [STARsoloFastq.summary],
+      align_features = [STARsoloFastq.align_features],
+      umipercell = [STARsoloFastq.umipercell],
       input_id = input_id,
       counting_mode = counting_mode,
       star_merge_docker_path = docker_prefix + star_merge_docker,
@@ -272,15 +252,15 @@ workflow Optimus {
   if (count_exons  && counting_mode=="sn_rna") {
     call StarAlign.MergeStarOutput as MergeStarOutputsExons {
       input:
-        barcodes = STARsoloFastq.barcodes_sn_rna,
-        features = STARsoloFastq.features_sn_rna,
-        matrix = STARsoloFastq.matrix_sn_rna,
-        cell_reads = STARsoloFastq.cell_reads_sn_rna,
+        barcodes = [STARsoloFastq.barcodes_sn_rna],
+        features = [STARsoloFastq.features_sn_rna],
+        matrix = [STARsoloFastq.matrix_sn_rna],
+        cell_reads = [STARsoloFastq.cell_reads_sn_rna],
         input_id = input_id,
         counting_mode = "sc_rna",
-        summary = STARsoloFastq.summary_sn_rna,
-        align_features = STARsoloFastq.align_features_sn_rna,
-        umipercell = STARsoloFastq.umipercell_sn_rna,
+        summary = [STARsoloFastq.summary_sn_rna],
+        align_features = [STARsoloFastq.align_features_sn_rna],
+        umipercell = [STARsoloFastq.umipercell_sn_rna],
         star_merge_docker_path = docker_prefix + star_merge_docker,
         gex_nhash_id = gex_nhash_id     
     }
@@ -351,7 +331,7 @@ workflow Optimus {
     # version of this pipeline
     String pipeline_version_out = pipeline_version
     File genomic_reference_version = ReferenceCheck.genomic_ref_version
-    File bam = MergeBam.output_bam
+    File bam = STARsoloFastq.bam_output
     File matrix = MergeStarOutputs.sparse_counts
     File matrix_row_index = MergeStarOutputs.row_index
     File matrix_col_index = MergeStarOutputs.col_index
@@ -363,12 +343,11 @@ workflow Optimus {
     File? mtx_files = MergeStarOutputs.mtx_files
     File? filtered_mtx_files = MergeStarOutputs.filtered_mtx_files
 
-    Array[File?] multimappers_EM_matrix = STARsoloFastq.multimappers_EM_matrix
-    Array[File?] multimappers_Uniform_matrix = STARsoloFastq.multimappers_Uniform_matrix
-    Array[File?] multimappers_Rescue_matrix = STARsoloFastq.multimappers_Rescue_matrix
-    Array[File?] multimappers_PropUnique_matrix = STARsoloFastq.multimappers_PropUnique_matrix
+    File? multimappers_EM_matrix = STARsoloFastq.multimappers_EM_matrix
+    File? multimappers_Uniform_matrix = STARsoloFastq.multimappers_Uniform_matrix
+    File? multimappers_Rescue_matrix = STARsoloFastq.multimappers_Rescue_matrix
+    File? multimappers_PropUnique_matrix = STARsoloFastq.multimappers_PropUnique_matrix
 
-
     # h5ad
     File h5ad_output_file = final_h5ad_output
 

pipelines/skylab/paired_tag/PairedTag.wdl

@@ -178,10 +178,10 @@ workflow PairedTag {
         File? cell_calls_gex = Optimus.cell_calls
         File h5ad_output_file_gex = Optimus.h5ad_output_file
         File? library_metrics = Optimus.library_metrics
-        Array[File?] multimappers_EM_matrix = Optimus.multimappers_EM_matrix
-        Array[File?] multimappers_Uniform_matrix = Optimus.multimappers_Uniform_matrix
-        Array[File?] multimappers_Rescue_matrix = Optimus.multimappers_Rescue_matrix
-        Array[File?] multimappers_PropUnique_matrix = Optimus.multimappers_PropUnique_matrix
+        File? multimappers_EM_matrix = Optimus.multimappers_EM_matrix
+        File? multimappers_Uniform_matrix = Optimus.multimappers_Uniform_matrix
+        File? multimappers_Rescue_matrix = Optimus.multimappers_Rescue_matrix
+        File? multimappers_PropUnique_matrix = Optimus.multimappers_PropUnique_matrix
         File? cell_barcodes_csv = Optimus.cell_barcodes_csv
         File? checkpoint_file = Optimus.checkpoint_file
         Array[File]? h5_array = Optimus.h5_array

tasks/skylab/StarAlign.wdl

@@ -224,21 +224,23 @@ task STARsoloFastq {
     String output_bam_basename
     Boolean? count_exons
     String? soloMultiMappers
+    String soloCBmatchWLtype = "1MM_multi" #"1MM_multi_Nbase_pseudocounts"
 
     # runtime values
     String samtools_star_docker_path
-    Int machine_mem_mb = 64000
-    Int cpu = 8
-   # by default request non preemptible machine to make sure the slow star alignment step completes
-    Int preemptible = 3
 
     # if slide_tags true set disk to 1000 otherwise dynamic allocation based on input size
     # dynamic allocation multiplies input size by 2.2 to account for output bam file + 20% overhead, add size of reference.
     Boolean is_slidetags
-    Int disk = if is_slidetags then 1000 else 
-    ceil(size(tar_star_reference, "Gi") * 3) + 
-    ceil(size(r1_fastq, "Gi") * 20) + 
-    ceil(size(r2_fastq, "Gi") * 20)
+
+    # runtime values
+    String cpu_platform = "Intel Ice Lake"
+    Int mem_size = 512
+    Int machine_mem_mb = 512000
+    Int cpu = 128
+    Int disk = 2000
+    # by default request non preemptible machine to make sure the slow star alignment step completes
+    Int preemptible = 1
   }
 
   meta {
@@ -314,7 +316,7 @@ task STARsoloFastq {
         echo Error: unknown counting mode: "$counting_mode". Should be either sn_rna or sc_rna.
         exit 1;
     fi
-
+    # RAM limit 33195969137
     STAR \
         --soloType Droplet \
         --soloStrand ~{star_strand_mode} \
@@ -327,11 +329,12 @@ task STARsoloFastq {
         --soloFeatures $COUNTING_MODE \
         --clipAdapterType CellRanger4 \
         --outFilterScoreMin 30  \
-        --soloCBmatchWLtype 1MM_multi \
+        --soloCBmatchWLtype ~{soloCBmatchWLtype} \
         --soloUMIdedup 1MM_CR \
         --outSAMtype BAM SortedByCoordinate \
         --outSAMattributes UB UR UY CR CB CY NH GX GN sF cN \
         --soloBarcodeReadLength 0 \
+        --limitBAMsortRAM 33195969137 \
         --soloCellReadStats Standard \
         ~{"--soloMultiMappers " + soloMultiMappers} \
         --soloUMIfiltering MultiGeneUMI_CR \
@@ -340,9 +343,7 @@ task STARsoloFastq {
     # validate the bam with samtools quickcheck
     samtools quickcheck -v Aligned.sortedByCoord.out.bam
 
-
     echo "UMI LEN " $UMILen
-
     touch barcodes_sn_rna.tsv
     touch features_sn_rna.tsv
     touch matrix_sn_rna.mtx
@@ -351,7 +352,6 @@ task STARsoloFastq {
     touch Summary_sn_rna.csv
     touch UMIperCellSorted_sn_rna.txt
 
-
     if [[ "~{counting_mode}" == "sc_rna" ]]
     then
       SoloDirectory="Solo.out/Gene/raw"
@@ -425,10 +425,11 @@ task STARsoloFastq {
 
   runtime {
     docker: samtools_star_docker_path
-    memory: "~{machine_mem_mb} MiB"
-    disks: "local-disk ~{disk} HDD"
+    memory: "~{mem_size} GiB"
+    disks: "local-disk ~{disk} SSD"
     disk: disk + " GB" # TES
     cpu: cpu
+    cpuPlatform: cpu_platform
     preemptible: preemptible
   }