Merge pull request #20 from uclahs-cds/nkwang-update-Rscripts

Update Rscripts
uclahs-cds · Aug 30, 2024 · dda927b · dda927b
2 parents 5c2bc28 + 9418d9c
commit dda927b
Show file tree

Hide file tree

Showing 15 changed files with 446 additions and 500 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -15,6 +15,8 @@ This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.htm
 - Add workflow for SV caller (Delly2)
 - Add pipeline diagram
 - Add reverse liftover (GRCh38 -> GRCh37) for SNV branch
+- Add reverse liftover (GRCh38 -> GRCh37) for SV branch
+- Add optional `target_threshold` and `target_specificity` parameters
 
 ### Changed
 

diff --git a/README.md b/README.md
@@ -51,10 +51,11 @@ If you are using the UCLA Azure cluster, please use the [submission script](http
 - For SNVs, convert variant coordinates using the `BCFtools` LiftOver plugin with UCSC chain files.
 - For SVs, convert variant breakpoint coordinates using custom R script with UCSC chain files and `rtracklayer` and `GenomicRanges` R packages.
 
-### 2. Variant annotation
+### 2. Variant annotation*
 
 - For SNVs, add dbSNP, GENCODE, and HGNC annotations using GATK's Funcotator. Add trinucleotide context and RepeatMasker intervals with `bedtools`.
 - For SVs, annotate variants with population allele frequency from the gnomAD-SV v4 database.
+- *Variant annotation occurs prior to LiftOver when converting from GRCh38 -> GRCh37
 
 ### 3. Predict variant stability
 
@@ -98,6 +99,8 @@ input:
 
 | Optional Parameter          | Type                                                                                      | Default                      | Description                                                                                                                                                                                                                                                                                                                                                                           |
 | --------------------------- | ----------------------------------------------------------------------------------------- | ---------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| `target_threshold`          | numeric                                                                                   | `""`             | Target Stability Score threshold for variant filtering: [0, 1] |
+| `target_specificity`          | numeric                                                                                   | `""`             | Target specificity based on whole genome validation set for variant filtering: [0, 1] |
 | `work_dir`                  | path                                                                                      | `/scratch/$SLURM_JOB_ID`     | Path of working directory for Nextflow. When included in the sample config file, Nextflow intermediate files and logs will be saved to this directory. With `ucla_cds`, the default is `/scratch` and should only be changed for testing/development. Changing this directory to `/hot` or `/tmp` can lead to high server latency and potential disk space limitations, respectively. |
 | `save_intermediate_files`   | boolean                                                                                   | false                        | If set, save output files from intermediate pipeline processes.                                                                                                                                                                                                                                                                                                                       |
 | `min_cpus`                  | int                                                                                       | 1                            | Minimum number of CPUs that can be assigned to each process.                                                                                                                                                                                                                                                                                                                          |
@@ -117,13 +120,13 @@ The docker images in the following table are generally defined like `docker_imag
 
 * Change `params.docker_container_registry`. This will affect all of the images (except for GATK).
 * Change `params.<tool>_version`. This will pull a different version of the same image from the registry.
-* Change `params.docker_image_<tool>`. This will explicitly set the image to use, ignoring `docker_container_registry` and `<tool>_version`, and thus requires that the docker tag be explicitly set (e.g. `broadinstitute/gatk:4.2.4.1`).
+* Change `params.docker_image_<tool>`. This will explicitly set the image to use, ignoring `docker_container_registry` and `<tool>_version`, and thus requires that the docker tag be explicitly set (e.g. `broadinstitute/gatk:4.4.0.0`).
 
 | Tool Parameter           | Version Parameter    | Default                                                      | Notes                                                               |
 | ------------------------ | -------------------- | ------------------------------------------------------------ | ------------------------------------------------------------------- |
 | `docker_image_bcftools`  | `bcftools_version`   | `ghcr.io/uclahs-cds/bcftools-score:1.20_score-1.20-20240505` | This image must have both BCFtools and the score plugins available. |
 | `docker_image_bedtools`  | `bedtools_version`   | `ghcr.io/uclahs-cds/bedtools:2.31.0`                         |                                                                     |
-| `docker_image_gatk`      | `gatk_version`       | `broadinstitute/gatk:4.2.4.1`                                |                                                                     |
+| `docker_image_gatk`      | `gatk_version`       | `broadinstitute/gatk:4.4.0.0`                                |                                                                     |
 | `docker_image_pipeval`   | `pipeval_version`    | `ghcr.io/uclahs-cds/pipeval:5.0.0-rc.3`                      |                                                                     |
 | `docker_image_samtools`  | `samtools_version`   | `ghcr.io/uclahs-cds/samtools:1.20`                           |                                                                     |
 | `doker_image_stablelift` | `stablelift_version` | `ghcr.io/uclahs-cds/stablelift:FIXME`                        | This image is built and maintained via this repository.             |
@@ -191,7 +194,7 @@ Please see list of [Contributors](https://github.com/uclahs-cds/pipeline-StableL
 
 pipeline-StableLift is licensed under the GNU General Public License version 2. See the file LICENSE for the terms of the GNU GPL license.
 
-StableLift is a machine learning approach designed to predict variant stability across reference genome builds, supplementing LiftOver coordinate conversion and increasing portability of variant calls.
+StableLift is a machine learning approach designed to predict variant stability across reference genome builds, supplementing LiftOver coordinate conversion to increase the portability of variant calls.
 
 Copyright (C) 2024 University of California Los Angeles ("Boutros Lab") All rights reserved.
 

diff --git a/config/custom_schema_types.config b/config/custom_schema_types.config
@@ -43,7 +43,29 @@ custom_schema_types {
         }
     }
 
+    /**
+    * Check that the input is numeric in the appropriate range.
+    */
+    ranged_number = { Map options, String name, Map properties ->
+        if (!(properties.containsKey('min') && properties['min'] in Number)) {
+            throw new Exception('`min` parameter misconfigured - must be a Number.')
+        }
+
+        if (!(properties.containsKey('max') && properties['max'] in Number)) {
+            throw new Exception('`max` parameter misconfigured - must be a Number.')
+        }
+
+        if (!(options[name] in Number)) {
+            throw new Exception("${name} must be a Number, not ${options[name].getClass()}")
+        }
+
+        if (options[name] < properties.min || properties.max < options[name]) {
+            throw new Exception("${name}=${options[name]} is not in range [${properties.min}, ${properties.max}]")
+        }
+    }
+
     types = [
-        'FuncotatorDataSource': custom_schema_types.check_funcotator_data_source
+        'FuncotatorDataSource': custom_schema_types.check_funcotator_data_source,
+        'RangedNumber': custom_schema_types.ranged_number
     ]
 }
diff --git a/config/schema.yaml b/config/schema.yaml
@@ -99,3 +99,23 @@ input:
       mode: 'r'
       required: true
       help: 'Input dataset supplied by input yaml'
+
+target_threshold:
+  type: 'RangedNumber'
+  required: false
+  min: 0
+  max: 1
+  help: >-
+    Optional parameter specifying target Stability Score threshold for variant
+    filtering Default behavior without `target_threshold` or
+    `target_specificity` specified uses threshold maximizing F1-score in whole
+    genome validation set'.
+
+target_specificity:
+  type: 'RangedNumber'
+  required: false
+  min: 0
+  max: 1
+  help: >-
+    Optional parameter specifying target specificity for variant filtering
+    based on whole genome validation set. Overrides `target_threshold`.
diff --git a/config/template.config b/config/template.config
@@ -5,17 +5,28 @@ includeConfig "${projectDir}/config/default.config"
 includeConfig "${projectDir}/config/methods.config"
 includeConfig "${projectDir}/nextflow.config"
 
-
 // Inputs/parameters of the pipeline
 params {
     // input/output locations
-    output_dir = 'where/to/save/outputs/'
+    output_dir = "where/to/save/outputs/"
 
-    // Choices: ["Mutect2", "HaplotypeCaller"]
+    // Choices: ["HaplotypeCaller", "Mutect2", "Strelka2", "SomaticSniper", "Muse2", "Delly2"]
     variant_caller = "Mutect2"
 
+    // Path to pre-trained random forest model
     rf_model = ""
 
+    // Optional parameter specifying target Stability Score threshold for
+    // variant filtering Default behavior without `target_threshold` or
+    // `target_specificity` specified uses threshold maximizing F1-score in
+    // whole genome validation set. Must be in the range [0.0, 1.0].
+    // target_threshold = 0.5
+
+    // Optional parameter specifying target specificity for variant filtering
+    // based on whole genome validation set Overrides `target_threshold`. Must
+    // be in the range [0.0, 1.0],
+    // target_specificity = 0.5
+
     // Reference files
     funcotator_data {
         data_source = "/hot/ref/tool-specific-input/Funcotator/somatic/funcotator_dataSources.v1.7.20200521s"

diff --git a/docs/pipeline.mmd b/docs/pipeline.mmd
@@ -64,7 +64,7 @@ flowchart TD
         --> bcftools_annotate2["`bcftools annotate*Trinucleotide*`"]:::bcftools
     end
 
-    blocknote["`**Note:** Annotation is performed before Liftover when lifting backward`"]
+    blocknote["`**Note:** Annotation is performed prior to LiftOver when converting from GRCh38 -> GRCh37`"]
 
     bcftools_liftover ---> gatk_func
     bcftools_annotate2 --> r_extract_snv[extract-VCF-features.R]:::R
@@ -79,7 +79,7 @@ flowchart TD
   joinpaths ---> r_predict_stability
 
   subgraph Predict Stability ["`&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;**Predict Stability**`"]
-    r_predict_stability[predict-liftover-stability.R]:::R
+    r_predict_stability[predict-variant-stability.R]:::R
     --> bcftools_annotate3["`bcftools annotate*Stability*`"]:::bcftools
 
     rf_model([rf_model]):::input .-> r_predict_stability

diff --git a/main.nf b/main.nf
@@ -42,6 +42,8 @@ log.info """\
         chain_file: ${params.chain_file}
         repeat_bed: ${params.repeat_bed}
 
+        header_contigs: ${params.getOrDefault('header_contigs', null)}
+
         funcotator_data:
             data_source:       ${params.funcotator_data.data_source}
             src_reference_id:  ${params.funcotator_data.src_reference_id}
@@ -149,10 +151,10 @@ workflow {
         // Take the SV branch
         workflow_extract_sv_annotations(
             validated_vcf_tuple,
+            input_ch_src_sequence,
             Channel.value(params.header_contigs),
             Channel.value(params.gnomad_rds),
-            Channel.value(params.chain_file),
-            Channel.value(params.variant_caller)
+            Channel.value(params.chain_file)
         )
 
         workflow_extract_sv_annotations.out.liftover_vcf.set { liftover_vcf }

diff --git a/module/predict_stability.nf b/module/predict_stability.nf
@@ -18,12 +18,17 @@ process predict_stability_StableLift {
     tuple val(sample_id), path("stability.tsv"), emit: stability_tsv
 
     script:
+    spec_arg = (params.getOrDefault('target_specificity', null) != null) ? "--specificity \"${params.get('target_specificity')}\"" : ""
+    thresh_arg = (params.getOrDefault('target_threshold', null) != null) ? "--threshold \"${params.get('target_threshold')}\"" : ""
+
     """
-    Rscript "${moduleDir}/scripts/predict-liftover-stability.R" \
+    Rscript "${moduleDir}/scripts/predict-variant-stability.R" \
+        --variant-caller "${variant_caller}" \
         --features-dt "${features_rds}" \
         --rf-model "${rf_model}" \
-        --variant-caller "${variant_caller}" \
-        --output-tsv "stability.tsv"
+        --output-tsv "stability.tsv" \
+        ${spec_arg} \
+        ${thresh_arg}
     """
 
     stub:
@@ -43,7 +48,6 @@ process run_apply_stability_annotations {
     input:
     tuple val(sample_id),
         path(annotated_vcf, stageAs: 'inputs/*'),
-        // FIXME Should there be an annotated_vcf_tbi?
         path(stability_tsv, stageAs: 'inputs/*'),
         path(stability_tsv_tbi, stageAs: 'inputs/*')