diff --git a/.dockstore.yml b/.dockstore.yml
index 9ab1966238..08449df04d 100644
--- a/.dockstore.yml
+++ b/.dockstore.yml
@@ -83,6 +83,14 @@ workflows:
     subclass: WDL
     primaryDescriptorPath: /pipelines/broad/arrays/imputation/Imputation.wdl
 
+  - name: ImputationBeagle
+    subclass: WDL
+    primaryDescriptorPath: /pipelines/broad/arrays/imputation_beagle/ImputationBeagle.wdl
+
+  - name: ArrayImputationQuotaConsumed
+    subclass: WDL
+    primaryDescriptorPath: /pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.wdl
+
   - name: RNAWithUMIsPipeline
     subclass: WDL
     primaryDescriptorPath: /pipelines/broad/rna_seq/RNAWithUMIsPipeline.wdl
@@ -155,6 +163,10 @@ workflows:
     subclass: WDL
     primaryDescriptorPath: /verification/test-wdls/TestImputation.wdl
 
+  - name: TestImputationBeagle
+    subclass: WDL
+    primaryDescriptorPath: /verification/test-wdls/TestImputationBeagle.wdl
+
   - name: TestJointGenotyping
     subclass: WDL
     primaryDescriptorPath: /verification/test-wdls/TestJointGenotyping.wdl
diff --git a/.github/workflows/test_imputation_beagle.yml b/.github/workflows/test_imputation_beagle.yml
new file mode 100644
index 0000000000..f9a627f02f
--- /dev/null
+++ b/.github/workflows/test_imputation_beagle.yml
@@ -0,0 +1,75 @@
+name: Test ImputationBeagle
+
+# Controls when the workflow will run
+on:
+  pull_request:
+    branches: [ "develop", "staging", "master" ]
+    # Only run if files in these paths changed:
+    ####################################
+    # SET PIPELINE SPECIFIC PATHS HERE #
+    ####################################
+    paths:
+      - 'pipelines/broad/arrays/imputation_beagle/**'
+      - 'structs/imputation/ImputationBeagleStructs.wdl'
+      - 'tasks/broad/ImputationTasks.wdl'
+      - 'tasks/broad/ImputationBeagleTasks.wdl'
+      - 'verification/VerifyImputationBeagle.wdl'
+      - 'verification/test-wdls/TestImputationBeagle.wdl'
+      - 'tasks/broad/Utilities.wdl'
+      - 'tasks/broad/TerraCopyFilesFromCloudToCloud.wdl'
+      - '.github/workflows/test_imputation_beagle.yml'
+      - '.github/workflows/warp_test_workflow.yml'
+
+
+  # Allows you to run this workflow manually from the Actions tab
+  workflow_dispatch:
+    inputs:
+      useCallCache:
+        description: 'Use call cache (default: true)'
+        required: false
+        default: "true"
+      updateTruth:
+        description: 'Update truth files (default: false)'
+        required: false
+        default: "false"
+      testType:
+        description: 'Specify the type of test (Plumbing or Scientific)'
+        required: false
+        type: choice
+        options:
+          - Plumbing
+          - Scientific
+      truthBranch:
+        description: 'Specify the branch for truth files (default: master)'
+        required: false
+        default: "master"
+
+env:
+  # pipeline configuration
+  PIPELINE_NAME: TestImputationBeagle
+  DOCKSTORE_PIPELINE_NAME: ImputationBeagle
+  PIPELINE_DIR: "pipelines/broad/arrays/imputation_beagle"
+
+  # workspace configuration
+  TESTING_WORKSPACE: WARP Tests
+  WORKSPACE_NAMESPACE: warp-pipelines
+
+  # service account configuration
+  SA_JSON_B64: ${{ secrets.PDT_TESTER_SA_B64 }}
+  USER: pdt-tester@warp-pipeline-dev.iam.gserviceaccount.com
+
+
+jobs:
+  TestImputationBeagle:
+    uses: ./.github/workflows/warp_test_workflow.yml
+    with:
+      pipeline_name: TestImputationBeagle
+      dockstore_pipeline_name: ImputationBeagle
+      pipeline_dir: pipelines/broad/arrays/imputation_beagle
+      use_call_cache: ${{ github.event.inputs.useCallCache || 'true' }}
+      update_truth: ${{ github.event.inputs.updateTruth || 'false' }}
+      test_type: ${{ github.event.inputs.testType }}
+      truth_branch: ${{ github.event.inputs.truthBranch || 'master' }}
+    secrets:
+      PDT_TESTER_SA_B64: ${{ secrets.PDT_TESTER_SA_B64 }}
+      DOCKSTORE_TOKEN: ${{ secrets.DOCKSTORE_TOKEN }}
diff --git a/pipeline_versions.txt b/pipeline_versions.txt
index 87e5e74d3b..bc8fa78585 100644
--- a/pipeline_versions.txt
+++ b/pipeline_versions.txt
@@ -1,7 +1,9 @@
 Pipeline Name	Version	Date of Last Commit
 Arrays	 2.6.30	2024-11-04 
 ValidateChip	 1.16.7	2024-11-04 
-Imputation	 1.1.15	2024-11-04 
+ArrayImputationQuotaConsumed	 1.0.0	2025-02-24 
+ImputationBeagle	 1.0.0	2025-02-26 
+Imputation	 1.1.16	2025-02-24 
 MultiSampleArrays	 1.6.2	2024-08-02 
 WholeGenomeReprocessing	 3.3.3	2024-11-04 
 ExomeReprocessing	 3.3.3	2024-11-04 
@@ -9,7 +11,7 @@ CramToUnmappedBams	 1.1.3	2024-08-02
 ExternalWholeGenomeReprocessing	 2.3.3	2024-11-04 
 ExternalExomeReprocessing	 3.3.3	2024-11-04 
 BroadInternalArrays	 1.1.14	2024-11-04 
-BroadInternalImputation	 1.1.14	2024-11-04 
+BroadInternalImputation	 1.1.15	2025-02-24 
 BroadInternalRNAWithUMIs	 1.0.36	2024-11-04 
 BroadInternalUltimaGenomics	 1.1.3	2024-12-05 
 RNAWithUMIsPipeline	 1.0.18	2024-11-04 
diff --git a/pipelines/broad/arrays/imputation/Imputation.changelog.md b/pipelines/broad/arrays/imputation/Imputation.changelog.md
index 52765e4ec1..5030cf3f05 100644
--- a/pipelines/broad/arrays/imputation/Imputation.changelog.md
+++ b/pipelines/broad/arrays/imputation/Imputation.changelog.md
@@ -1,3 +1,8 @@
+# 1.1.16
+2025-02-24 (Date of Last Commit)
+
+* Updated runtime parameters in some ImputationTasks, and added an explicit definition of a vcf_index.
+
 # 1.1.15
 2024-11-04 (Date of Last Commit)
 
diff --git a/pipelines/broad/arrays/imputation/Imputation.wdl b/pipelines/broad/arrays/imputation/Imputation.wdl
index 4a44ba4ac5..3466169b64 100644
--- a/pipelines/broad/arrays/imputation/Imputation.wdl
+++ b/pipelines/broad/arrays/imputation/Imputation.wdl
@@ -6,7 +6,7 @@ import "../../../../tasks/broad/Utilities.wdl" as utils
 
 workflow Imputation {
 
-  String pipeline_version = "1.1.15"
+  String pipeline_version = "1.1.16"
 
   input {
     Int chunkLength = 25000000
@@ -242,6 +242,7 @@ workflow Imputation {
       call tasks.SelectVariantsByIds {
         input:
           vcf = SetIdsVcfToImpute.output_vcf,
+          vcf_index = SetIdsVcfToImpute.output_vcf_index,
           ids = FindSitesUniqueToFileTwoOnly.missing_sites,
           basename = "imputed_sites_to_recover"
       }
diff --git a/pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.changelog.md b/pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.changelog.md
new file mode 100644
index 0000000000..978888b711
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.changelog.md
@@ -0,0 +1,4 @@
+# 1.0.0
+2025-02-24 (Date of Last Commit)
+
+* Initial release of pipeline to calculate the number of samples, i.e. quota used by an imputation service that uses ImputationBeagle.wdl.
diff --git a/pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.wdl b/pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.wdl
new file mode 100644
index 0000000000..a4cd6e8d09
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/ArrayImputationQuotaConsumed.wdl
@@ -0,0 +1,29 @@
+version 1.0
+
+import "../../../../tasks/broad/ImputationTasks.wdl" as tasks
+
+workflow QuotaConsumed {
+    String pipeline_version = "1.0.0"
+
+    input {
+        Int chunkLength = 25000000
+        Int chunkOverlaps = 5000000
+
+        File multi_sample_vcf
+
+        File ref_dict
+        Array[String] contigs
+        String reference_panel_path_prefix
+        String genetic_maps_path
+        String output_basename
+    }
+
+    call tasks.CountSamples {
+        input:
+            vcf = multi_sample_vcf
+    }
+
+    output {
+        Int quota_consumed = CountSamples.nSamples
+    }
+}
diff --git a/pipelines/broad/arrays/imputation_beagle/ImputationBeagle.changelog.md b/pipelines/broad/arrays/imputation_beagle/ImputationBeagle.changelog.md
new file mode 100644
index 0000000000..ddc7604697
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/ImputationBeagle.changelog.md
@@ -0,0 +1,5 @@
+# 1.0.0
+2025-02-26 (Date of Last Commit)
+
+* Initial public release of the ImputationBeagle pipeline.
+  * The ImputationBeagle pipeline imputes missing genotypes from a multi-sample VCF using a large genomic reference panel. It is based on the Michigan Imputation Server pipeline but uses the Beagle imputation tool instead of minimac. Overall, the pipeline filters, phases, and performs imputation on a multi-sample VCF. It outputs the imputed VCF.
diff --git a/pipelines/broad/arrays/imputation_beagle/ImputationBeagle.wdl b/pipelines/broad/arrays/imputation_beagle/ImputationBeagle.wdl
new file mode 100644
index 0000000000..64d058b965
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/ImputationBeagle.wdl
@@ -0,0 +1,226 @@
+version 1.0
+
+import "../../../../structs/imputation/ImputationBeagleStructs.wdl" as structs
+import "../../../../tasks/broad/ImputationTasks.wdl" as tasks
+import "../../../../tasks/broad/ImputationBeagleTasks.wdl" as beagleTasks
+
+workflow ImputationBeagle {
+
+  String pipeline_version = "1.0.0"
+
+  input {
+    Int chunkLength = 25000000
+    Int chunkOverlaps = 2000000 # this is the padding that will be added to the beginning and end of each chunk to reduce edge effects
+
+    File multi_sample_vcf
+
+    File ref_dict # for reheadering / adding contig lengths in the header of the ouptut VCF, and calculating contig lengths
+    Array[String] contigs
+    String reference_panel_path_prefix # path + file prefix to the bucket where the reference panel files are stored for all contigs
+    String genetic_maps_path # path to the bucket where genetic maps are stored for all contigs
+    String output_basename # the basename for intermediate and output files
+
+    # file extensions used to find reference panel files
+    String bed_suffix = ".bed"
+    String bref3_suffix = ".bref3"
+
+    String gatk_docker = "broadinstitute/gatk:4.6.0.0"
+    String ubuntu_docker = "ubuntu:20.04"
+
+    Int? error_count_override
+  }
+
+  call tasks.CountSamples {
+    input:
+      vcf = multi_sample_vcf
+  }
+
+  call beagleTasks.CreateVcfIndex {
+    input:
+      vcf_input = multi_sample_vcf,
+      gatk_docker = gatk_docker
+  }
+
+  Float chunkLengthFloat = chunkLength
+
+  scatter (contig in contigs) {
+    # these are specific to hg38 - contig is format 'chr1'
+    String reference_basename = reference_panel_path_prefix + "." + contig
+    String genetic_map_filename = genetic_maps_path + "plink." + contig + ".GRCh38.withchr.map"
+
+    ReferencePanelContig referencePanelContig = {
+      "bed": reference_basename + bed_suffix,
+      "bref3": reference_basename  + bref3_suffix,
+      "contig": contig,
+      "genetic_map": genetic_map_filename
+    }
+
+
+    call tasks.CalculateChromosomeLength {
+      input:
+        ref_dict = ref_dict,
+        chrom = referencePanelContig.contig,
+        ubuntu_docker = ubuntu_docker
+    }
+
+    Int num_chunks = ceil(CalculateChromosomeLength.chrom_length / chunkLengthFloat)
+
+    scatter (i in range(num_chunks)) {
+      String chunk_contig = referencePanelContig.contig
+
+      Int start = (i * chunkLength) + 1
+      Int startWithOverlaps = if (start - chunkOverlaps < 1) then 1 else start - chunkOverlaps
+      Int end = if (CalculateChromosomeLength.chrom_length < ((i + 1) * chunkLength)) then CalculateChromosomeLength.chrom_length else ((i + 1) * chunkLength)
+      Int endWithOverlaps = if (CalculateChromosomeLength.chrom_length < end + chunkOverlaps) then CalculateChromosomeLength.chrom_length else end + chunkOverlaps
+      String chunk_basename = referencePanelContig.contig + "_chunk_" + i
+
+      # generate the chunked vcf file that will be used for imputation, including overlaps
+      call tasks.GenerateChunk {
+        input:
+          vcf = CreateVcfIndex.vcf,
+          vcf_index = CreateVcfIndex.vcf_index,
+          start = startWithOverlaps,
+          end = endWithOverlaps,
+          chrom = referencePanelContig.contig,
+          basename = chunk_basename,
+          gatk_docker = gatk_docker
+      }
+
+      call beagleTasks.CountVariantsInChunks {
+        input:
+          vcf = GenerateChunk.output_vcf,
+          vcf_index = GenerateChunk.output_vcf_index,
+          panel_bed_file = referencePanelContig.bed,
+          gatk_docker = gatk_docker
+      }
+
+      call beagleTasks.CheckChunks {
+        input:
+          var_in_original = CountVariantsInChunks.var_in_original,
+          var_also_in_reference = CountVariantsInChunks.var_also_in_reference
+      }
+    }
+
+    Array[File] chunkedVcfsWithOverlapsForImputation = GenerateChunk.output_vcf
+
+    call tasks.StoreChunksInfo as StoreContigLevelChunksInfo {
+      input:
+        chroms = chunk_contig,
+        starts = start,
+        ends = end,
+        vars_in_array = CountVariantsInChunks.var_in_original,
+        vars_in_panel = CountVariantsInChunks.var_also_in_reference,
+        valids = CheckChunks.valid,
+        basename = output_basename
+    }
+
+    # if any chunk for any chromosome fail CheckChunks, then we will not impute run any task in the next scatter,
+    # namely phasing and imputing which would be the most costly to throw away
+    Int n_failed_chunks_int = select_first([error_count_override, read_int(StoreContigLevelChunksInfo.n_failed_chunks)])
+    call beagleTasks.ErrorWithMessageIfErrorCountNotZero as FailQCNChunks {
+      input:
+        errorCount = n_failed_chunks_int,
+        message = "contig " + referencePanelContig.contig + " had " + n_failed_chunks_int + " failing chunks"
+    }
+
+    scatter (i in range(num_chunks)) {
+      String chunk_basename_imputed = referencePanelContig.contig + "_chunk_" + i + "_imputed"
+
+      # max amount of cpus you can ask for is 96 so at a max of 10k samples we can only ask for 9 cpu a sample.
+      # these values are based on trying to optimize for pre-emptibility using a 400k sample reference panel
+      # and up to a 10k sample input vcf
+      Int beagle_cpu = if (CountSamples.nSamples <= 1000) then 8 else floor(CountSamples.nSamples / 1000) * 9
+      Int beagle_phase_memory_in_gb = if (CountSamples.nSamples <= 1000) then 22 else ceil(beagle_cpu * 1.5)
+      Int beagle_impute_memory_in_gb = if (CountSamples.nSamples <= 1000) then 30 else ceil(beagle_cpu * 4.3)
+
+      call beagleTasks.Phase {
+        input:
+          dataset_vcf = chunkedVcfsWithOverlapsForImputation[i],
+          ref_panel_bref3 = referencePanelContig.bref3,
+          chrom = referencePanelContig.contig,
+          basename = chunk_basename_imputed,
+          genetic_map_file = referencePanelContig.genetic_map,
+          start = start[i],
+          end = end[i],
+          cpu = beagle_cpu,
+          memory_mb = beagle_phase_memory_in_gb * 1024,
+          for_dependency = FailQCNChunks.done
+      }
+
+      call beagleTasks.Impute {
+        input:
+          dataset_vcf = Phase.vcf,
+          ref_panel_bref3 = referencePanelContig.bref3,
+          chrom = referencePanelContig.contig,
+          basename = chunk_basename_imputed,
+          genetic_map_file = referencePanelContig.genetic_map,
+          start = start[i],
+          end = end[i],
+          cpu = beagle_cpu,
+          memory_mb = beagle_impute_memory_in_gb * 1024
+      }
+
+      call beagleTasks.CreateVcfIndex as IndexImputedBeagle {
+        input:
+          vcf_input = Impute.vcf,
+          gatk_docker = gatk_docker
+      }
+
+      call tasks.UpdateHeader {
+        input:
+          vcf = IndexImputedBeagle.vcf,
+          vcf_index = IndexImputedBeagle.vcf_index,
+          ref_dict = ref_dict,
+          basename = chunk_basename_imputed,
+          disable_sequence_dictionary_validation = false,
+          gatk_docker = gatk_docker
+      }
+
+      call tasks.SeparateMultiallelics {
+        input:
+          original_vcf = UpdateHeader.output_vcf,
+          original_vcf_index = UpdateHeader.output_vcf_index,
+          output_basename = chunk_basename_imputed
+      }
+
+      call tasks.RemoveSymbolicAlleles {
+        input:
+          original_vcf = SeparateMultiallelics.output_vcf,
+          original_vcf_index = SeparateMultiallelics.output_vcf_index,
+          output_basename = chunk_basename_imputed,
+          gatk_docker = gatk_docker
+      }
+    }
+
+    Array[File] chromosome_vcfs = select_all(RemoveSymbolicAlleles.output_vcf)
+  }
+
+  call tasks.GatherVcfs {
+    input:
+      input_vcfs = flatten(chromosome_vcfs),
+      output_vcf_basename = output_basename + ".imputed",
+      gatk_docker = gatk_docker
+  }
+
+  call tasks.StoreChunksInfo {
+    input:
+      chroms = flatten(chunk_contig),
+      starts = flatten(start),
+      ends = flatten(end),
+      vars_in_array = flatten(CountVariantsInChunks.var_in_original),
+      vars_in_panel = flatten(CountVariantsInChunks.var_also_in_reference),
+      valids = flatten(CheckChunks.valid),
+      basename = output_basename
+  }
+  
+  output {
+    File imputed_multi_sample_vcf = GatherVcfs.output_vcf
+    File imputed_multi_sample_vcf_index = GatherVcfs.output_vcf_index
+    File chunks_info = StoreChunksInfo.chunks_info
+  }
+
+  meta {
+    allowNestedInputs: true
+  }
+
+}
diff --git a/pipelines/broad/arrays/imputation_beagle/README.md b/pipelines/broad/arrays/imputation_beagle/README.md
new file mode 100644
index 0000000000..754e416b5a
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/README.md
@@ -0,0 +1,7 @@
+### ImputationBeagle summary
+
+The ImputationBeagle pipeline imputes missing genotypes from a multi-sample VCF using the [Beagle imputation tool](https://faculty.washington.edu/browning/beagle/beagle.html) and a large genomic reference panel. Overall, the pipeline filters, phases, and performs imputation on a multi-sample VCF. This pipeline was created for use by the All of Us/AnVIL Imputation Service.
+
+### ArrayImputationQuotaConsumed summary
+
+The ArrayImputationQuotaConsumed pipeline is used by the All of Us/AnVIL Imputation Service and calculates the number of samples in the input multi-sample VCF, which is the metric used by the service for ImputationBeagle pipeline quota.
diff --git a/pipelines/broad/arrays/imputation_beagle/test_inputs/Plumbing/NA12878_x10_hg38_arrays.json b/pipelines/broad/arrays/imputation_beagle/test_inputs/Plumbing/NA12878_x10_hg38_arrays.json
new file mode 100644
index 0000000000..bdf5a00597
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/test_inputs/Plumbing/NA12878_x10_hg38_arrays.json
@@ -0,0 +1,8 @@
+{
+    "ImputationBeagle.multi_sample_vcf": "gs://broad-gotc-test-storage/imputation_beagle/scientific/vcfs/NA12878_10_duplicate.merged.cleaned.vcf.gz",
+    "ImputationBeagle.ref_dict": "gs://gcp-public-data--broad-references/hg38/v0/Homo_sapiens_assembly38.dict",
+    "ImputationBeagle.reference_panel_path_prefix": "gs://broad-gotc-test-storage/imputation_beagle/scientific/1000G_HGDP_no_singletons_reference_panel/hgdp.tgp.gwaspy.AN_added.bcf.ac2",
+    "ImputationBeagle.contigs": ["chr21","chr22"],
+    "ImputationBeagle.genetic_maps_path": "gs://broad-gotc-test-storage/imputation_beagle/scientific/plink-genetic-maps/",
+    "ImputationBeagle.output_basename": "plumbing_test"
+}
diff --git a/pipelines/broad/arrays/imputation_beagle/test_inputs/Scientific/NA12878_x10_hg38_arrays.json b/pipelines/broad/arrays/imputation_beagle/test_inputs/Scientific/NA12878_x10_hg38_arrays.json
new file mode 100644
index 0000000000..4263609e29
--- /dev/null
+++ b/pipelines/broad/arrays/imputation_beagle/test_inputs/Scientific/NA12878_x10_hg38_arrays.json
@@ -0,0 +1,8 @@
+{
+    "ImputationBeagle.multi_sample_vcf": "gs://broad-gotc-test-storage/imputation_beagle/scientific/vcfs/NA12878_10_duplicate.merged.cleaned.vcf.gz",
+    "ImputationBeagle.ref_dict": "gs://gcp-public-data--broad-references/hg38/v0/Homo_sapiens_assembly38.dict",
+    "ImputationBeagle.reference_panel_path_prefix": "gs://broad-gotc-test-storage/imputation_beagle/scientific/1000G_HGDP_no_singletons_reference_panel/hgdp.tgp.gwaspy.AN_added.bcf.ac2",
+    "ImputationBeagle.contigs": ["chr1","chr2","chr3","chr4","chr5","chr6","chr7","chr8","chr9","chr10","chr11","chr12","chr13","chr14","chr15","chr16","chr17","chr18","chr19","chr20","chr21","chr22"],
+    "ImputationBeagle.genetic_maps_path": "gs://broad-gotc-test-storage/imputation_beagle/scientific/plink-genetic-maps/",
+    "ImputationBeagle.output_basename": "scientific_test"
+}
diff --git a/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.changelog.md b/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.changelog.md
index a0930046d7..e4f328a7fb 100644
--- a/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.changelog.md
+++ b/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.changelog.md
@@ -1,3 +1,8 @@
+# 1.1.15
+2025-02-24 (Date of Last Commit)
+
+* Updated runtime parameters in some ImputationTasks, and added an explicit definition of a vcf_index.
+
 # 1.1.14
 2024-11-04 (Date of Last Commit)
 
diff --git a/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.wdl b/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.wdl
index 525ce85e00..27e16fa28e 100644
--- a/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.wdl
+++ b/pipelines/broad/internal/arrays/imputation/BroadInternalImputation.wdl
@@ -9,7 +9,7 @@ workflow BroadInternalImputation {
         description: "Push outputs of Imputation.wdl to TDR dataset table ImputationOutputsTable and split out Imputation arrays into ImputationWideOutputsTable."
         allowNestedInputs: true
     }
-    String pipeline_version = "1.1.14"
+    String pipeline_version = "1.1.15"
     
     input {
         # inputs to wrapper task 
diff --git a/structs/imputation/ImputationBeagleStructs.wdl b/structs/imputation/ImputationBeagleStructs.wdl
new file mode 100644
index 0000000000..6aaaaa061a
--- /dev/null
+++ b/structs/imputation/ImputationBeagleStructs.wdl
@@ -0,0 +1,8 @@
+version 1.0
+
+struct ReferencePanelContig {
+  File bed
+  File bref3
+  String contig
+  File genetic_map
+}
diff --git a/tasks/broad/ImputationBeagleTasks.wdl b/tasks/broad/ImputationBeagleTasks.wdl
new file mode 100644
index 0000000000..af4363e600
--- /dev/null
+++ b/tasks/broad/ImputationBeagleTasks.wdl
@@ -0,0 +1,217 @@
+version 1.0
+
+task CountVariantsInChunks {
+  input {
+    File vcf
+    File vcf_index
+    File panel_bed_file
+
+    String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.5.0.0"
+    Int cpu = 1
+    Int memory_mb = 16000
+    Int disk_size_gb = 2 * ceil(size([vcf, vcf_index, panel_bed_file], "GiB")) + 10
+  }
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
+
+  command <<<
+    set -e -o pipefail
+
+    ln -sf ~{vcf} input.vcf.gz
+    ln -sf ~{vcf_index} input.vcf.gz.tbi
+
+    gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" CountVariants -V input.vcf.gz | tail -n 1 > var_in_original
+    bedtools intersect -a ~{vcf} -b ~{panel_bed_file} | wc -l > var_also_in_reference
+  >>>
+
+  output {
+    Int var_in_original = read_int("var_in_original")
+    Int var_also_in_reference = read_int("var_also_in_reference")
+  }
+  runtime {
+    docker: gatk_docker
+    disks: "local-disk ${disk_size_gb} HDD"
+    memory: "${memory_mb} MiB"
+    cpu: cpu
+    preemptible: 3
+  }
+}
+
+task CheckChunks {
+  input {
+    Int var_in_original
+    Int var_also_in_reference
+
+    String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
+    Int cpu = 1
+    Int memory_mb = 4000
+  }
+  command <<<
+    set -e -o pipefail
+
+    if [ $(( ~{var_also_in_reference} * 2 - ~{var_in_original})) -gt 0 ] && [ ~{var_also_in_reference} -gt 3 ]; then
+      echo true > valid_file.txt
+    else
+      echo false > valid_file.txt
+    fi
+  >>>
+  output {
+    Boolean valid = read_boolean("valid_file.txt")
+  }
+  runtime {
+    docker: bcftools_docker
+    disks: "local-disk 10 HDD"
+    memory: "${memory_mb} MiB"
+    cpu: cpu
+    preemptible: 3
+  }
+}
+
+task Phase {
+  input {
+    File dataset_vcf
+    File ref_panel_bref3
+    File genetic_map_file
+    String basename
+    String chrom
+    Int start
+    Int end
+
+    String beagle_docker = "us.gcr.io/broad-gotc-prod/imputation-beagle:1.0.0-17Dec24.224-1740423035"
+    Int cpu = 8                    # This parameter is used as the nthreads input to Beagle which is part of how we make it determinstic.  Changing this value may change the output generated by the tool
+    Int memory_mb = 32000          # value depends on chunk size, the number of samples in ref and target panel, and whether imputation is performed
+    Int xmx_mb = memory_mb - 5000             # I suggest setting this parameter to be 85-90% of the memory_mb parameter
+    Int disk_size_gb = ceil(3 * size([dataset_vcf, ref_panel_bref3], "GiB")) + 10         # value may need to be adjusted
+
+    Boolean for_dependency         # used for task dependency management
+  }
+
+  command <<<
+    set -e -o pipefail
+
+    java -ea -Xmx~{xmx_mb}m \
+    -jar /usr/gitc/beagle.17Dec24.224.jar \
+    gt=~{dataset_vcf} \
+    ref=~{ref_panel_bref3} \
+    map=~{genetic_map_file} \
+    out=phased_~{basename} \
+    chrom=~{chrom}:~{start}-~{end} \
+    impute=false \
+    nthreads=~{cpu} \
+    seed=-99999
+
+  >>>
+  output {
+    File vcf = "phased_~{basename}.vcf.gz"
+    File log = "phased_~{basename}.log"
+  }
+  runtime {
+    docker: beagle_docker
+    disks: "local-disk ${disk_size_gb} HDD"
+    memory: "${memory_mb} MiB"
+    cpu: cpu
+    preemptible: 3
+  }
+}
+
+task Impute {
+  input {
+    File dataset_vcf
+    File ref_panel_bref3
+    File genetic_map_file
+    String basename
+    String chrom
+    Int start
+    Int end
+
+    String beagle_docker = "us.gcr.io/broad-gotc-prod/imputation-beagle:1.0.0-17Dec24.224-1740423035"
+    Int cpu = 8                    # This parameter is used as the nthreads input to Beagle which is part of how we make it determinstic.  Changing this value may change the output generated by the tool
+    Int memory_mb = 32000          # value depends on chunk size, the number of samples in ref and target panel, and whether imputation is performed
+    Int xmx_mb = memory_mb - 5000             # I suggest setting this parameter to be 85-90% of the memory_mb parameter
+    Int disk_size_gb = ceil(3 * size([dataset_vcf, ref_panel_bref3], "GiB")) + 10         # value may need to be adjusted
+  }
+
+  command <<<
+    set -e -o pipefail
+
+    java -ea -Xmx~{xmx_mb}m \
+    -jar /usr/gitc/beagle.17Dec24.224.jar \
+    gt=~{dataset_vcf} \
+    ref=~{ref_panel_bref3} \
+    map=~{genetic_map_file} \
+    out=imputed_~{basename} \
+    chrom=~{chrom}:~{start}-~{end} \
+    impute=true \
+    nthreads=~{cpu} \
+    seed=-99999
+
+  >>>
+  output {
+    File vcf = "imputed_~{basename}.vcf.gz"
+    File log = "imputed_~{basename}.log"
+  }
+  runtime {
+    docker: beagle_docker
+    disks: "local-disk ${disk_size_gb} HDD"
+    memory: "${memory_mb} MiB"
+    cpu: cpu
+    preemptible: 3
+  }
+}
+
+task ErrorWithMessageIfErrorCountNotZero {
+  input {
+    Int errorCount
+    String message
+  }
+  command <<<
+    if [[ ~{errorCount} -gt 0 ]]; then
+      >&2 echo "Error: ~{message}"
+      exit 1
+    else
+      exit 0
+    fi
+  >>>
+
+  runtime {
+    docker: "ubuntu:20.04"
+    preemptible: 3
+  }
+  output {
+    Boolean done = true
+  }
+}
+
+task CreateVcfIndex {
+  input {
+    File vcf_input
+
+    Int disk_size_gb = ceil(1.2*size(vcf_input, "GiB")) + 10
+    Int cpu = 1
+    Int memory_mb = 6000
+    String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.5.0.0"
+  }
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
+
+  String vcf_basename = basename(vcf_input)
+
+  command {
+    set -e -o pipefail
+
+    ln -sf ~{vcf_input} ~{vcf_basename}
+
+    bcftools index -t ~{vcf_basename}
+  }
+  runtime {
+    docker: gatk_docker
+    disks: "local-disk ${disk_size_gb} HDD"
+    memory: "${memory_mb} MiB"
+    cpu: cpu
+    preemptible: 3
+  }
+  output {
+    File vcf = "~{vcf_basename}"
+    File vcf_index = "~{vcf_basename}.tbi"
+  }
+}
diff --git a/tasks/broad/ImputationTasks.wdl b/tasks/broad/ImputationTasks.wdl
index 793ae119b2..1a89954535 100644
--- a/tasks/broad/ImputationTasks.wdl
+++ b/tasks/broad/ImputationTasks.wdl
@@ -12,6 +12,8 @@ task CalculateChromosomeLength {
   }
 
   command {
+    set -e -o pipefail
+
     grep -P "SN:~{chrom}\t" ~{ref_dict} | sed 's/.*LN://' | sed 's/\t.*//'
   }
   runtime {
@@ -19,6 +21,7 @@ task CalculateChromosomeLength {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     Int chrom_length = read_int(stdout())
@@ -37,6 +40,8 @@ task GetMissingContigList {
   }
 
   command <<<
+    set -e -o pipefail
+
     grep "@SQ" ~{ref_dict} | sed 's/.*SN://' | sed 's/\t.*//' > contigs.txt
     awk 'NR==FNR{arr[$0];next} !($0 in arr)' ~{included_contigs} contigs.txt > missing_contigs.txt
   >>>
@@ -62,13 +67,13 @@ task GenerateChunk {
     File vcf
     File vcf_index
 
-    Int disk_size_gb = ceil(2*size(vcf, "GiB")) + 50 # not sure how big the disk size needs to be since we aren't downloading the entire VCF here
+    Int disk_size_gb = ceil(2*size(vcf, "GiB")) + 10
     Int cpu = 1
     Int memory_mb = 8000
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
 
   command {
     gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" \
@@ -114,17 +119,17 @@ task CountVariantsInChunks {
 
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
     Int cpu = 1
-    Int memory_mb = 4000
+    Int memory_mb = 6000
     Int disk_size_gb = 2 * ceil(size([vcf, vcf_index, panel_vcf, panel_vcf_index], "GiB")) + 20
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
 
   command <<<
     set -e -o pipefail
 
     echo $(gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" CountVariants -V ~{vcf}  | sed 's/Tool returned://') > var_in_original
-    echo $(gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" CountVariants -V ~{vcf} -L ~{panel_vcf}  | sed 's/Tool returned://') > var_in_reference
+    echo $(gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" CountVariants -V ~{vcf} -L ~{panel_vcf} | sed 's/Tool returned://') > var_in_reference
   >>>
   output {
     Int var_in_original = read_int("var_in_original")
@@ -147,7 +152,7 @@ task CheckChunks {
     Int var_in_original
     Int var_in_reference
 
-    Int disk_size_gb = ceil(2*size([vcf, vcf_index, panel_vcf, panel_vcf_index], "GiB"))
+    Int disk_size_gb = ceil(2*size([vcf, vcf_index, panel_vcf, panel_vcf_index], "GiB")) + 10
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 4000
@@ -191,7 +196,7 @@ task PhaseVariantsEagle {
     String eagle_docker = "us.gcr.io/broad-gotc-prod/imputation-eagle:1.0.0-2.4-1690199702"
     Int cpu = 8
     Int memory_mb = 32000
-    Int disk_size_gb = ceil(3 * size([dataset_bcf, reference_panel_bcf, dataset_bcf_index, reference_panel_bcf_index], "GiB")) + 50
+    Int disk_size_gb = ceil(3 * size([dataset_bcf, reference_panel_bcf, dataset_bcf_index, reference_panel_bcf_index], "GiB")) + 10
   }
   command <<<
     /usr/gitc/eagle  \
@@ -269,10 +274,10 @@ task GatherVcfs {
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
     Int cpu = 1
     Int memory_mb = 16000
-    Int disk_size_gb = ceil(3*size(input_vcfs, "GiB"))
+    Int disk_size_gb = ceil(3*size(input_vcfs, "GiB")) + 10
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
 
   command <<<
     set -e -o pipefail
@@ -285,7 +290,6 @@ task GatherVcfs {
 
     gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" \
     IndexFeatureFile -I ~{output_vcf_basename}.vcf.gz
-
   >>>
   runtime {
     docker: gatk_docker
@@ -304,6 +308,8 @@ task ReplaceHeader {
     File vcf_to_replace_header
     File vcf_with_new_header
 
+    Int cpu = 1
+    Int memory_mb = 6000
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
   }
 
@@ -321,6 +327,9 @@ task ReplaceHeader {
   runtime {
     docker: bcftools_docker
     disks: "local-disk ${disk_size_gb} HDD"
+    memory: "${memory_mb} MiB"
+    cpu: cpu
+    preemptible: 3
   }
 
   output {
@@ -334,30 +343,32 @@ task UpdateHeader {
     File vcf_index
     File ref_dict
     String basename
+    Boolean disable_sequence_dictionary_validation = true
 
     Int disk_size_gb = ceil(4*(size(vcf, "GiB") + size(vcf_index, "GiB"))) + 20
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
     Int cpu = 1
-    Int memory_mb = 8000
+    Int memory_mb = 6000
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
+  String disable_sequence_dict_validation_flag = if disable_sequence_dictionary_validation then "--disable-sequence-dictionary-validation" else ""
 
   command <<<
-
     ## update the header of the merged vcf
     gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" \
     UpdateVCFSequenceDictionary \
     --source-dictionary ~{ref_dict} \
     --output ~{basename}.vcf.gz \
     --replace -V ~{vcf} \
-    --disable-sequence-dictionary-validation
+    ~{disable_sequence_dict_validation_flag}
   >>>
   runtime {
     docker: gatk_docker
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     File output_vcf = "~{basename}.vcf.gz"
@@ -376,8 +387,8 @@ task RemoveSymbolicAlleles {
     Int cpu = 1
     Int memory_mb = 4000
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
 
   command {
     gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" \
@@ -392,6 +403,7 @@ task RemoveSymbolicAlleles {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
 }
 
@@ -401,7 +413,7 @@ task SeparateMultiallelics {
     File original_vcf_index
     String output_basename
 
-    Int disk_size_gb =  ceil(2*(size(original_vcf, "GiB") + size(original_vcf_index, "GiB")))
+    Int disk_size_gb =  ceil(2*(size(original_vcf, "GiB") + size(original_vcf_index, "GiB"))) + 10
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 4000
@@ -421,6 +433,7 @@ task SeparateMultiallelics {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
 }
 
@@ -435,7 +448,7 @@ task OptionalQCSites {
     String bcftools_vcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 16000
-    Int disk_size_gb = ceil(2*(size(input_vcf, "GiB") + size(input_vcf_index, "GiB")))
+    Int disk_size_gb = ceil(2*(size(input_vcf, "GiB") + size(input_vcf_index, "GiB"))) + 10
 
   }
   Float max_missing = select_first([optional_qc_max_missing, 0.05])
@@ -443,8 +456,11 @@ task OptionalQCSites {
   command <<<
     set -e -o pipefail
 
+    ln -sf ~{input_vcf} input.vcf.gz
+    ln -sf ~{input_vcf_index} input.vcf.gz.tbi
+
     # site missing rate < 5% ; hwe p > 1e-6
-    vcftools --gzvcf ~{input_vcf}  --max-missing ~{max_missing} --hwe ~{hwe} --recode -c | bgzip -c > ~{output_vcf_basename}.vcf.gz
+    vcftools --gzvcf input.vcf.gz --max-missing ~{max_missing} --hwe ~{hwe} --recode -c | bgzip -c > ~{output_vcf_basename}.vcf.gz
     bcftools index -t ~{output_vcf_basename}.vcf.gz # Note: this is necessary because vcftools doesn't have a way to output a zipped vcf, nor a way to index one (hence needing to use bcf).
   >>>
   runtime {
@@ -472,6 +488,7 @@ task MergeSingleSampleVcfs {
   }
   command <<<
     set -e -o pipefail
+
     # Move the index file next to the vcf with the corresponding name
 
     declare -a VCFS=(~{sep=' ' input_vcfs})
@@ -507,9 +524,12 @@ task CountSamples {
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 3000
-    Int disk_size_gb = 100 + ceil(size(vcf, "GiB"))
+    Int disk_size_gb = ceil(size(vcf, "GiB")) + 10
   }
+
   command <<<
+    set -e -o pipefail
+
     bcftools query -l ~{vcf} | wc -l
   >>>
   runtime {
@@ -517,6 +537,7 @@ task CountSamples {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     Int nSamples = read_int(stdout())
@@ -532,7 +553,7 @@ task AggregateImputationQCMetrics {
     String rtidyverse_docker = "rocker/tidyverse:4.1.0"
     Int cpu = 1
     Int memory_mb = 2000
-    Int disk_size_gb = 100 + ceil(size(infoFile, "GiB"))
+    Int disk_size_gb = ceil(size(infoFile, "GiB")) + 10
   }
   command <<<
     Rscript -<< "EOF"
@@ -560,7 +581,7 @@ task AggregateImputationQCMetrics {
     disks : "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
-    preemptible : 3
+    preemptible: 3
   }
   output {
     File aggregated_metrics = "~{basename}_aggregated_imputation_metrics.tsv"
@@ -600,7 +621,7 @@ task StoreChunksInfo {
     disks : "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
-    preemptible : 3
+    preemptible: 3
   }
   output {
     File chunks_info = "~{basename}_chunk_info.tsv"
@@ -617,7 +638,7 @@ task MergeImputationQCMetrics {
     String rtidyverse_docker = "rocker/tidyverse:4.1.0"
     Int cpu = 1
     Int memory_mb = 2000
-    Int disk_size_gb = 100 + ceil(size(metrics, "GiB"))
+    Int disk_size_gb = ceil(size(metrics, "GiB")) + 10
   }
   command <<<
     Rscript -<< "EOF"
@@ -638,7 +659,7 @@ task MergeImputationQCMetrics {
     disks : "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
-    preemptible : 3
+    preemptible: 3
   }
   output {
     File aggregated_metrics = "~{basename}_aggregated_imputation_metrics.tsv"
@@ -656,13 +677,13 @@ task SubsetVcfToRegion {
     Int end
     Boolean exclude_filtered = false
 
-    Int disk_size_gb = ceil(2*size(vcf, "GiB")) + 50 # not sure how big the disk size needs to be since we aren't downloading the entire VCF here
+    Int disk_size_gb = ceil(2*size(vcf, "GiB")) + 10
     Int cpu = 1
     Int memory_mb = 8000
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
 
   command {
     gatk --java-options "-Xms~{command_mem}m -Xmx~{max_heap}m" \
@@ -705,10 +726,11 @@ task SetIDs {
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 4000
-    Int disk_size_gb = 100 + ceil(2.2 * size(vcf, "GiB"))
+    Int disk_size_gb = ceil(2.2 * size(vcf, "GiB")) + 10
   }
   command <<<
     set -e -o pipefail
+
     bcftools annotate ~{vcf} --set-id '%CHROM\:%POS\:%REF\:%FIRST_ALT' -Oz -o ~{output_basename}.vcf.gz
     bcftools index -t ~{output_basename}.vcf.gz
   >>>
@@ -717,6 +739,7 @@ task SetIDs {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     File output_vcf = "~{output_basename}.vcf.gz"
@@ -729,7 +752,7 @@ task ExtractIDs {
     File vcf
     String output_basename
 
-    Int disk_size_gb = 2*ceil(size(vcf, "GiB")) + 100
+    Int disk_size_gb = 2*ceil(size(vcf, "GiB")) + 10
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 4000
@@ -745,28 +768,34 @@ task ExtractIDs {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
 }
 
 task SelectVariantsByIds {
   input {
     File vcf
+    File vcf_index
     File ids
     String basename
 
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
     Int cpu = 1
     Int memory_mb = 16000
-    Int disk_size_gb = ceil(1.2*size(vcf, "GiB")) + 100
+    Int disk_size_gb = ceil(1.2*size(vcf, "GiB")) + 10
   }
   parameter_meta {
     vcf: {
       description: "vcf",
       localization_optional: true
     }
+    vcf_index: {
+      description: "vcf",
+      localization_optional: true
+    }
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 2000
+  Int max_heap = memory_mb - 1500
 
   command <<<
     set -e -o pipefail
@@ -780,6 +809,7 @@ task SelectVariantsByIds {
     disks: "local-disk ${disk_size_gb} SSD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     File output_vcf = "~{basename}.vcf.gz"
@@ -795,7 +825,7 @@ task RemoveAnnotations {
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
     Int memory_mb = 3000
-    Int disk_size_gb = ceil(2.2*size(vcf, "GiB")) + 100
+    Int disk_size_gb = ceil(2.2*size(vcf, "GiB")) + 10
   }
   command <<<
     set -e -o pipefail
@@ -808,6 +838,7 @@ task RemoveAnnotations {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     File output_vcf = "~{basename}.vcf.gz"
@@ -823,10 +854,10 @@ task InterleaveVariants {
     String gatk_docker = "us.gcr.io/broad-gatk/gatk:4.6.1.0"
     Int cpu = 1
     Int memory_mb = 16000
-    Int disk_size_gb = ceil(3.2*size(vcfs, "GiB")) + 100
+    Int disk_size_gb = ceil(3.2*size(vcfs, "GiB")) + 10
   }
-  Int command_mem = memory_mb - 1000
-  Int max_heap = memory_mb - 500
+  Int command_mem = memory_mb - 1500
+  Int max_heap = memory_mb - 1000
 
   command <<<
     set -e -o pipefail
@@ -839,6 +870,7 @@ task InterleaveVariants {
     disks: "local-disk ${disk_size_gb} SSD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     File output_vcf = "~{basename}.vcf.gz"
@@ -854,9 +886,11 @@ task FindSitesUniqueToFileTwoOnly {
     String ubuntu_docker = "ubuntu:20.04"
     Int cpu = 1
     Int memory_mb = 4000
-    Int disk_size_gb = ceil(size(file1, "GiB") + 2*size(file2, "GiB")) + 100
+    Int disk_size_gb = ceil(size(file1, "GiB") + 2*size(file2, "GiB")) + 10
   }
   command <<<
+    set -e -o pipefail
+
     comm -13 <(sort ~{file1} | uniq) <(sort ~{file2} | uniq) > missing_sites.ids
   >>>
   runtime {
@@ -864,6 +898,7 @@ task FindSitesUniqueToFileTwoOnly {
     disks: "local-disk ${disk_size_gb} HDD"
     memory: "${memory_mb} MiB"
     cpu: cpu
+    preemptible: 3
   }
   output {
     File missing_sites = "missing_sites.ids"
@@ -877,10 +912,10 @@ task SplitMultiSampleVcf {
 
     String bcftools_docker = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
     Int cpu = 1
-    Int memory_mb = 8000
+    Int memory_mb = 6000
 
     # This calculation is explained in https://github.com/broadinstitute/warp/pull/937
-    Int disk_size_gb = ceil(21*nSamples*size(multiSampleVcf, "GiB")/(nSamples+20)) + 100
+    Int disk_size_gb = ceil(21*nSamples*size(multiSampleVcf, "GiB")/(nSamples+20)) + 10
   }
   command <<<
     set -e -o pipefail
@@ -901,4 +936,4 @@ task SplitMultiSampleVcf {
     Array[File] single_sample_vcfs = glob("out_dir/*.vcf.gz")
     Array[File] single_sample_vcf_indices = glob("out_dir/*.vcf.gz.tbi")
   }
-}
\ No newline at end of file
+}
diff --git a/verification/VerifyImputationBeagle.wdl b/verification/VerifyImputationBeagle.wdl
new file mode 100644
index 0000000000..e99f1767c1
--- /dev/null
+++ b/verification/VerifyImputationBeagle.wdl
@@ -0,0 +1,79 @@
+version 1.0
+
+import "../verification/VerifyTasks.wdl" as Tasks
+
+## Copyright Broad Institute, 2018
+##
+## This WDL script is designed to verify (compare) the outputs of an ArrayWf wdl.
+##
+##
+## Runtime parameters are optimized for Broad's Google Cloud Platform implementation.
+## For program versions, see docker containers.
+##
+## LICENSING :
+## This script is released under the WDL source code license (BSD-3) (see LICENSE in
+## https://github.com/broadinstitute/wdl). Note however that the programs it calls may
+## be subject to different licenses. Users are responsible for checking that they are
+## authorized to run all programs before running this script. Please see the docker
+## page at https://hub.docker.com/r/broadinstitute/genomes-in-the-cloud/ for detailed
+## licensing information pertaining to the included programs.
+
+workflow VerifyImputationBeagle {
+  input {
+    Array[File] truth_metrics
+    Array[File] test_metrics
+
+    File truth_vcf
+    File test_vcf
+    File test_vcf_index
+    File truth_vcf_index
+
+    Boolean? done
+  }
+
+  String bcftools_docker_tag = "us.gcr.io/broad-gotc-prod/imputation-bcf-vcf:1.0.7-1.10.2-0.1.16-1669908889"
+
+  scatter (idx in range(length(truth_metrics))) {
+    call CompareImputationMetrics {
+      input:
+        test_metrics = test_metrics[idx],
+        truth_metrics = truth_metrics[idx]
+    }
+  }
+
+  call Tasks.CompareVcfs as CompareOutputVcfs {
+    input:
+      file1 = truth_vcf,
+      file2 = test_vcf,
+      patternForLinesToExcludeFromComparison = "##" # ignore headers
+  }
+
+  output {
+  }
+  meta {
+    allowNestedInputs: true
+  }
+}
+
+task CompareImputationMetrics {
+  input {
+    File test_metrics
+    File truth_metrics
+  }
+  command <<<
+    set -eo pipefail
+    diff "~{test_metrics}" "~{truth_metrics}"
+
+    if [ $? -ne 0 ];
+    then
+      echo "Error: ${test_metrics} and ${truth_metrics}  differ"
+    fi
+  >>>
+
+  runtime {
+    docker: "ubuntu:20.04"
+    cpu: 1
+    memory: "3.75 GiB"
+    disks: "local-disk 10 HDD"
+  }
+}
diff --git a/verification/VerifyTasks.wdl b/verification/VerifyTasks.wdl
index 43bb9b4340..46ed24373c 100644
--- a/verification/VerifyTasks.wdl
+++ b/verification/VerifyTasks.wdl
@@ -10,7 +10,7 @@ task CompareVcfs {
   command {
     set -eo pipefail
 
-    if [ -z ~{patternForLinesToExcludeFromComparison} ]; then
+    if [ -z '~{patternForLinesToExcludeFromComparison}' ]; then
       diff <(gunzip -c -f ~{file1}) <(gunzip -c -f ~{file2})
     else
       echo "It's defined!"
diff --git a/verification/test-wdls/TestImputationBeagle.wdl b/verification/test-wdls/TestImputationBeagle.wdl
new file mode 100644
index 0000000000..2d56d858aa
--- /dev/null
+++ b/verification/test-wdls/TestImputationBeagle.wdl
@@ -0,0 +1,111 @@
+version 1.0
+
+
+import "../../pipelines/broad/arrays/imputation_beagle/ImputationBeagle.wdl" as ImputationBeagle
+import "../../verification/VerifyImputationBeagle.wdl" as VerifyImputationBeagle
+import "../../tasks/broad/Utilities.wdl" as Utilities
+import "../../tasks/broad/TerraCopyFilesFromCloudToCloud.wdl" as Copy
+
+workflow TestImputationBeagle {
+
+    input {
+      Int chunkLength = 25000000
+      Int chunkOverlaps = 5000000 # this is the padding that will be added to the beginning and end of each chunk to reduce edge effects
+      
+      File multi_sample_vcf
+      
+      File ref_dict # for reheadering / adding contig lengths in the header of the ouptut VCF, and calculating contig lengths
+      Array[String] contigs
+      String reference_panel_path_prefix # path + file prefix to the bucket where the reference panel files are stored for all contigs
+      String genetic_maps_path # path to the bucket where genetic maps are stored for all contigs
+      String output_basename # the basename for intermediate and output files
+
+      # These values will be determined and injected into the inputs by the scala test framework
+      String truth_path
+      String results_path
+      Boolean update_truth
+    }
+
+    meta {
+      allowNestedInputs: true
+    }
+  
+    call ImputationBeagle.ImputationBeagle {
+      input:
+        chunkLength = chunkLength,
+        chunkOverlaps = chunkOverlaps,
+        multi_sample_vcf = multi_sample_vcf,
+        ref_dict = ref_dict,
+        contigs = contigs,
+        reference_panel_path_prefix = reference_panel_path_prefix,
+        genetic_maps_path = genetic_maps_path,
+        output_basename = output_basename,
+    }
+
+    
+    # Collect all of the pipeline outputs into single Array[String]
+    Array[String] pipeline_outputs = flatten([
+                                    [ # File outputs
+                                    ImputationBeagle.imputed_multi_sample_vcf,
+                                    ImputationBeagle.imputed_multi_sample_vcf_index,
+                                    ]
+    ])
+
+    
+    # Collect all of the pipeline metrics into single Array[String]
+    Array[String] pipeline_metrics = flatten([
+                                    [ # File outputs
+                                    ImputationBeagle.chunks_info,
+                                    ]
+    ])
+
+    # Copy results of pipeline to test results bucket
+    call Copy.TerraCopyFilesFromCloudToCloud as CopyToTestResults {
+      input:
+        files_to_copy             = flatten([pipeline_outputs, pipeline_metrics]),
+        destination_cloud_path    = results_path
+    }
+  
+    # If updating truth then copy output to truth bucket
+    if (update_truth){
+      call Copy.TerraCopyFilesFromCloudToCloud as CopyToTruth {
+        input: 
+          files_to_copy             = flatten([pipeline_outputs, pipeline_metrics]),
+          destination_cloud_path    = truth_path
+      }
+    }
+
+    # This is achieved by passing each desired file/array[files] to GetValidationInputs
+    if (!update_truth){
+        call Utilities.GetValidationInputs as GetMetrics {
+          input:
+            input_files = pipeline_metrics,
+            results_path = results_path,
+            truth_path = truth_path
+        }
+        call Utilities.GetValidationInputs as GetVcf {
+          input:
+            input_file = ImputationBeagle.imputed_multi_sample_vcf,
+            results_path = results_path,
+            truth_path = truth_path
+        }
+        call Utilities.GetValidationInputs as GetVcfIndex {
+          input:
+            input_file = ImputationBeagle.imputed_multi_sample_vcf_index,
+            results_path = results_path,
+            truth_path = truth_path
+        }
+
+
+      call VerifyImputationBeagle.VerifyImputationBeagle as Verify {
+        input:
+          truth_metrics = GetMetrics.truth_files, 
+          test_metrics = GetMetrics.results_files,
+          truth_vcf = GetVcf.truth_file, 
+          test_vcf = GetVcf.results_file,
+          truth_vcf_index = GetVcfIndex.truth_file, 
+          test_vcf_index = GetVcfIndex.results_file,
+          done = CopyToTestResults.done
+      }
+    }
+}