docs update and runtime params

docs/workflows/standalone/dorado_basecalling.md

@@ -12,11 +12,11 @@ The Dorado Basecalling workflow is used to convert Oxford Nanopore `POD5` sequen
 
 ### Model Type Selection
 
-Users can choose between automatic or manual model selection using a configurable use_auto_model flag:
+Users can configure the basecalling model by setting the dorado_model input parameter:
 
-Automatic Model Selection: Automatically picks the best model ('sup', 'hac', or 'fast') based on the input file and user-defined model accuracy paramater.
-
-Manual Model Input: If the user disables automatic selection, a specific model path or model version must be provided.
+- Default Model: "sup" (super accuracy) is used unless overridden by the user.
+- Manual Model Input: Users can specify the full path or name of a specific model (e.g., [email protected]).
+- Automatic Detection: When set to sup, hac, or fast, Dorado will automatically select the appropriate model version if available.
 
 - **Model Type (sup):** (super accuracy) The most accurate model, recommended for critical applications requiring the highest basecall accuracy. It is the slowest of the three model types.
 - **Model Type (hac):** (High Accuracy) A balance between speed and accuracy, recommended for most users. Provides accurate results faster than `sup` but less accurate than `sup`.
@@ -42,10 +42,8 @@ Manual Model Input: If the user disables automatic selection, a specific model p
 | **Task** | **Variable** | **Type** | **Description** | **Default Value** | **Required** |
 |---|---|---|---|---|---|
 | Basecalling | **input_files** | Array[File] | Array of `POD5` files for basecalling | None | Yes |
-| Basecalling | **use_auto_model** | Boolean | Use automatic model selection (`sup`, `hac`, or `fast` based on model accuracy)| true | No |
-| Basecalling | **model_accuracy** | String | Desired model accuracy (`sup`, `hac`, `fast`) if using automatic selection | sup | No |
+| Basecalling | **dorado_model** | Boolean | Model accuracy or full model name (default: 'sup')| sup | No |
 | Basecalling | **fastq_file_name** | String | Prefix for naming output FASTQ files | None | Yes |
-| Basecalling | **dorado_model** | String | Model type (e.g., `[email protected]`) if manual input | None | Yes |
 | Basecalling | **kit_name** | String | Sequencing kit name used (e.g., `SQK-RPB114-24`). | None | Yes |
 | Basecalling | **cpu** | Int | Number of CPUs allocated | 8 | No |
 | Basecalling | **memory** | String | Amount of memory to allocate | 32GB | No |
@@ -59,6 +57,8 @@ Manual Model Input: If the user disables automatic selection, a specific model p
 ---
 
 ### Detailed Input Information
+
+- **dorado_model**: If set to 'sup', 'hac', or 'fast', the workflow will run with automatic model selection. If a full model name is provided, Dorado will use that model directly.
 - **fastq_file_name**: This will serve as a prefix for the output FASTQ files. For example, if you provide `project001`, the resulting files will be named `project001_barcodeXX.fastq.gz`.
 - **kit_name**: Ensure the correct kit name is provided, as it determines the barcoding and adapter trimming behavior.
 - **fastq_upload_path**: This is the folder path in Terra where the final FASTQ files will be transferred for further analysis. Ensure the path matches your Terra workspace configuration.

tasks/basecalling/task_dorado_basecall.wdl

@@ -5,6 +5,9 @@ task basecall {
     File input_file                    # Single POD5 file for scatter processing
     String dorado_model = "sup"         # Default model to 'sup', can be overridden with full model name see docs
     String kit_name                    # Sequencing kit name
+    Int disk_size = 100
+    Int memory = 32
+    Int cpu = 8
     String docker = "us-docker.pkg.dev/general-theiagen/staphb/dorado:0.8.0"
   }
 
@@ -41,8 +44,10 @@ task basecall {
 
   runtime {
     docker: docker
-    cpu: 8
-    memory: "32GB"
+    cpu: cpu
+    memory: "~{memory} GB"
+    disks:  "local-disk " + disk_size + " SSD"
+    disk: disk_size + " GB"
     gpuCount: 1
     gpuType: "nvidia-tesla-t4"  }
 }

tasks/basecalling/task_dorado_demux.wdl

@@ -5,6 +5,9 @@ task dorado_demux {
     Array[File] bam_files
     String kit_name
     String fastq_file_name
+    Int cpu = 4 
+    Int memory = 16
+    Int disk_size = 50
     String docker = "us-docker.pkg.dev/general-theiagen/staphb/dorado:0.8.0"
   }
 
@@ -83,10 +86,10 @@ task dorado_demux {
     Array[File] fastq_files = glob("~{fastq_file_name}_*.fastq.gz")
   }
 
-  runtime {
+   runtime {
     docker: docker
-    cpu: 4
-    memory: "16GB"
-    maxRetries: 0
+    cpu: cpu
+    memory: "~{memory} GB"
+    disks: "local-disk ~{disk_size} SSD"
   }
 }

tasks/basecalling/task_samtools_convert.wdl

@@ -3,6 +3,9 @@ version 1.0
 task samtools_convert {
   input {
     Array[File] sam_files
+    Int cpu = 4
+    Int memory = 16
+    Int disk_size = 50
     String docker = "us-docker.pkg.dev/general-theiagen/staphb/samtools:1.15"
   }
 
@@ -30,10 +33,10 @@ task samtools_convert {
     Array[File] bam_files = glob("output/bam/*.bam")
   }
 
-  runtime {
+ runtime {
     docker: docker
-    cpu: 4
-    memory: "16GB"
-    maxRetries: 0
+    cpu: cpu
+    memory: "~{memory} GB"
+    disks: "local-disk ~{disk_size} HDD"
   }
 }

workflows/utilities/wf_dorado_basecalling.wdl

@@ -13,7 +13,7 @@ workflow dorado_basecalling_workflow {
 
   input {
     Array[File] input_files
-    String dorado_model = "sup"  # Default to sup model, user can override with a full model name
+    String dorado_model = "sup" 
     String kit_name
     String new_table_name
     String fastq_upload_path
@@ -23,14 +23,20 @@ workflow dorado_basecalling_workflow {
     String terra_project
     String terra_workspace
     String fastq_file_name
+    Int cpu = 8                      
+    Int memory = 32                    
+    Int disk_size = 100               
   }
 
   scatter (file in input_files) {
     call basecall_task.basecall as basecall_step {
       input:
         input_file = file,
         dorado_model = dorado_model,
-        kit_name = kit_name
+        kit_name = kit_name,
+        cpu = cpu,
+        memory = memory,
+        disk_size = disk_size
     }
   }