Untested rest of workflow

README.md

@@ -1,54 +1,3 @@
-This is a template to use to extend the [Sunbeam pipeline](https://github.com/sunbeam-labs/sunbeam). There are three major parts to a Sunbeam extension: 
-
- - `sbx_sga_env.yml` specifies the extension's dependencies
- - `config.yml` contains configuration options that can be specified by the user when running an extension
- - `sbx_sga.rules` contains the rules (logic/commands run) of the extension
-
-## Creating an extension
-
-Any dependencies (available through conda) required for an extension should be listed in the `sbx_[your_extension_name]_env.yml` file. An example of how to format this is shown in the `sbx_sga_env.yml` file. These dependencies are automatically handled and installed by Snakemake through conda (see below for how to make sure your rule finds this env file), so the user doesn't have to worry about installing dependencies themselves. You can also specify specific software versions like so:
-
-    dependencies:
-      - python=3.7
-      - megahit<2
-
-Rarely, you may want to specify different environments for different rules within your extension, in the event that different rules have different (potentially conflicting) requirements.
-
-The `config.yml` contains parameters that the user might need to modify when running an extension. For example, if your downstream analysis is run differently depending on whether reads are paired- or single-end, it would probably be wise to include a `paired_end` parameter. Default values should be specified for each terminal key. As of Sunbeam v3.0, as long as the parameter config file is named `config.yml`, configuration options for installed extensions are automatically included in new config files generated by `sunbeam init` and `sunbeam config update`.
-
-Finally, `sbx_sga.rules` contains the actual logic for the extension, including required input and output files. A detailed discussion of Snakemake rule creation is beyond the scope of this tutorial, but definitely check out [the Snakemake tutorial](http://snakemake.readthedocs.io/en/stable/tutorial/basics.html) and any of the [extensions by sunbeam-labs](https://github.com/sunbeam-labs) for inspiration.
-
-For each rule that needs dependencies from your environment file, make sure to let snakemake know in the rule like this:
-
-    example_rule:
-        ...
-        conda:
-            "sbx_sga_env.yml"
-        ...
-
-The dependency .yml file can be named whatever you want, as long as you refer to it by the correct filename in whatever rule needs those dependencies. The path to the dependency .yml file is relative to the .rules file (which in most cases is in the same directory).
-
-## Additional extension components
-
-### .github/
-
-This directory contains CI workflows for GitHub to run automatically on PRs, including tests and linting. If the linter raises errors, you can fix them by running `snakefmt` on any snakemake files and `black` on any python files. The release workflow will build and push a docker image for each environment in the extension.
-
-### .tests/
-
-This directory contains tests, broken down into types such as end-to-end (e2e) and unit, as well as data for running these tests.
-
-### scripts/
-
-This directory contains scripts that can be run by rules. Use this for any rules that need to run python, R, etc code.
-
-### envs/*.Dockerfile
-
-The Dockerfiles provided with conda env specifications allow for containerized runs of sunbeam (meaning they use docker containers to run each rule rather than conda envs).
-
-(You can delete everything above this line)
------------------------------------------------------------------
-
 <img src="https://github.com/sunbeam-labs/sunbeam/blob/stable/docs/images/sunbeam_logo.gif" width=120, height=120 align="left" />
 
 # sbx_sga

config.yml

@@ -1,2 +1,4 @@
 sbx_sga:
-  example_rule_options: 'DUMMY STRING'
+  mash_ref: RefSeq88n.msh
+  checkm_ref: /mnt/isilon/marc_genomics/shared_data/CheckM2_database/uniref100.KO.1.dmnd
+  bakta_ref: /mnt/isilon/marc_genomics/shared_data/bakta_db_5/db/

envs/abritamr.yml

@@ -0,0 +1,7 @@
+name: sga_abritamr
+channels:
+  - bioconda
+  - conda-forge
+  - defaults
+dependencies:
+  - abritamr

envs/bakta.yml

@@ -0,0 +1,7 @@
+name: sga_bakta
+channels:
+  - bioconda
+  - conda-forge
+  - defaults
+dependencies:
+  - bakta

envs/checkm2.yml

@@ -0,0 +1,6 @@
+name: sga_checkm2
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - checkm2

envs/sbx_sga_env.yml → envs/mash.yml

@@ -1,6 +1,6 @@
-name: sbx_sga
+name: sga_mash
 channels:
   - conda-forge
   - bioconda
 dependencies:
-  - samtools
+  - mash

envs/mlst.yml

@@ -0,0 +1,6 @@
+name: sga_mlst
+channels:
+  - bioconda
+  - conda-forge
+dependencies:
+  - mlst

envs/quast.yml

@@ -0,0 +1,7 @@
+name: sga_quast
+channels:
+  - bioconda
+  - conda-forge
+  - defaults
+dependencies:
+  - quast

envs/shovill.yml

@@ -0,0 +1,6 @@
+name: sga_shovill
+channels:
+  - conda-forge
+  - bioconda
+dependencies:
+  - shovill

sbx_sga.smk

@@ -1,4 +1,4 @@
-def get_template_path() -> Path:
+def get_sga_path() -> Path:
     for fp in sys.path:
         if fp.split("/")[-1] == "sbx_sga":
             return Path(fp)
@@ -7,7 +7,8 @@ def get_template_path() -> Path:
     )
 
 
-SBX_TEMPLATE_VERSION = open(get_template_path() / "VERSION").read().strip()
+ISOLATE_FP = Cfg["all"]["output_fp"] / "isolate"
+SBX_SBA_VERSION = open(get_sga_path() / "VERSION").read().strip()
 
 try:
     BENCHMARK_FP
@@ -20,47 +21,173 @@ except NameError:
 
 
 localrules:
-    all_template,
+    all_sga,
 
 
-rule all_template:
+rule all_sga:
     input:
-        QC_FP / "mush" / "big_file.txt",
+        # QC
+        #expand(ISOLATE_FP / "fastqc" / "{sample}_{rp}_fastqc/fastqc_data.txt", rp=Pairs, sample=Samples),
+        expand(ISOLATE_FP / "mash" / "{sample}_sorted_winning.tab", sample=Samples),
+        # Assembly QC
+        ISOLATE_FP / "checkm" / "quality_report.tsv",
+        expand(ISOLATE_FP / "quast" / "{sample}" / "report.tsv", sample=Samples),
+        # Typing
+        ISOLATE_FP / "mlst" / "mlst_report.tsv",
+        # Annotation
+        expand(ISOLATE_FP / "bakta" / "{sample}" / "{sample}.txt", sample=Samples),
+        # AMR Profiling
+        expand(ISOLATE_FP / "abritamr" / "{sample}" / "amrfinder.out", sample=Samples),
 
 
-rule example_rule:
-    """Takes in cleaned .fastq.gz and mushes them all together into a file"""
+rule sga_mash:
     input:
-        expand(QC_FP / "cleaned" / "{sample}_{rp}.fastq.gz", sample=Samples, rp=Pairs),
+        reads=expand(QC_FP / "decontam" / "{{sample}}_{rp}.fastq.gz", rp=Pairs),
     output:
-        QC_FP / "mush" / "big_file1.txt",
+        agg=temp(ISOLATE_FP / "mash" / "{sample}.fastq"),
+        win=temp(ISOLATE_FP / "mash" / "{sample}_winning.tab"),
+        sort=ISOLATE_FP / "mash" / "{sample}_sorted_winning.tab",
+    params:
+        ref=Cfg["sbx_sga"]["mash_ref"],
     log:
-        LOG_FP / "example_rule.log",
+        LOG_FP / "sga_mash_{sample}.log",
     benchmark:
-        BENCHMARK_FP / "example_rule.tsv"
+        BENCHMARK_FP / "sga_mash_{sample}.tsv"
+    conda:
+        "envs/mash.yml"
+    shell:
+        """
+        zcat  {input.reads} > {output.agg}
+        mash screen -w -p 8 {params.ref} {output.agg} > {output.win} 2> {log}
+        sort -gr {output.win} > {output.sort} 2>> {log}
+        """
+
+
+rule sga_shovill:
+    input:
+        rp1=QC_FP / "decontam" / "{sample}_1.fastq.gz",
+        rp2=QC_FP / "decontam" / "{sample}_2.fastq.gz",
+    output:
+        contigs=ISOLATE_FP / "shovill" / "{sample}" / "contigs.fa",
+    log:
+        LOG_FP / "sga_shovill_{sample}.log",
+    benchmark:
+        BENCHMARK_FP / "sga_shovill_{sample}.tsv"
+    conda:
+        "envs/shovill.yml"
+    shell:
+        """
+        shovill --assembler skesa --outdir $(dirname {output.contigs}) --R1 {input.rp1}  --R2 {input.rp2} &> {log}
+        """
+
+
+### Assembly QC
+rule sga_checkm:
+    input:
+        contigs=expand(
+            ISOLATE_FP / "shovill" / "{sample}" / "contigs.fa", sample=Samples
+        ),
+    output:
+        quality_report=ISOLATE_FP / "checkm" / "quality_report.tsv",
     params:
-        opts=Cfg["sbx_sga"]["example_rule_options"],
+        ref=Cfg["sbx_sga"]["checkm_ref"],
+    log:
+        LOG_FP / "sga_checkm.log",
+    benchmark:
+        BENCHMARK_FP / "sga_checkm.tsv"
     conda:
-        "envs/sbx_sga_env.yml"
-    container:
-        f"docker://sunbeamlabs/sbx_sga:{SBX_TEMPLATE_VERSION}"
+        "envs/checkm2.yml"
     shell:
-        "cat {params.opts} {input} >> {output} 2> {log}"
+        """
+        checkm2 predict \\
+        -x fa \\
+        -i $(dirname {input.contigs[0]}) \\
+        -o $(dirname {output.quality_report}) \\
+        --database_path {params.ref} \\
+        &> {log}
+        """
 
 
-rule example_with_script:
-    """Take in big_file1 and then ignore it and write the results of `samtools --help` to the output using a python script"""
+rule sga_quast:
     input:
-        QC_FP / "mush" / "big_file1.txt",
+        contigs=ISOLATE_FP / "shovill" / "{sample}" / "contigs.fa",
     output:
-        QC_FP / "mush" / "big_file.txt",
+        quast_dir=ISOLATE_FP / "quast" / "{sample}" / "report.tsv",
     log:
-        LOG_FP / "example_with_script.log",
+        LOG_FP / "sga_quast_{sample}.log",
     benchmark:
-        BENCHMARK_FP / "example_with_script.tsv"
+        BENCHMARK_FP / "sga_quast_{sample}.tsv"
     conda:
-        "envs/sbx_sga_env.yml"
-    container:
-        f"docker://sunbeamlabs/sbx_sga:{SBX_TEMPLATE_VERSION}"
-    script:
-        "scripts/example_with_script.py"
+        "envs/quast.yml"
+    shell:
+        """
+        quast.py \\
+        -o $(dirname {output.quast_dir}) \\
+        {input.contigs} \\
+        &> {log}
+        """
+
+
+### Typing
+rule sga_mlst:
+    input:
+        contigs=expand(
+            ISOLATE_FP / "shovill" / "{sample}" / "contigs.fa", sample=Samples
+        ),
+    output:
+        mlst=ISOLATE_FP / "mlst" / "mlst_report.tsv",
+    log:
+        LOG_FP / "sga_mlst.log",
+    benchmark:
+        BENCHMARK_FP / "sga_mlst.tsv"
+    conda:
+        "envs/mlst.yml"
+    shell:
+        """
+        mlst {input.contigs} > {output.mlst} 2> {log}
+        """
+
+
+### Annotation
+rule sga_bakta:
+    input:
+        contigs=ISOLATE_FP / "shovill" / "{sample}" / "contigs.fa",
+    output:
+        bakta=ISOLATE_FP / "bakta" / "{sample}" / "{sample}.txt",
+    params:
+        ref=Cfg["sbx_sga"]["bakta_ref"],
+    log:
+        LOG_FP / "sga_bakta_{sample}.log",
+    benchmark:
+        BENCHMARK_FP / "sga_bakta_{sample}.tsv"
+    conda:
+        "envs/bakta.yml"
+    shell:
+        """
+        bakta --db {params.ref} \\
+        --output $(dirname {output.bakta}) \\
+        --prefix {wildcards.sample} \\
+        --skip-plot {input.contigs} \\
+        &> {log}
+        """
+
+
+### AMR Profiling
+rule sga_abritamr:
+    input:
+        contigs=ISOLATE_FP / "shovill" / "{sample}" / "contigs.fa",
+    output:
+        abritamr=ISOLATE_FP / "abritamr" / "{sample}" / "amrfinder.out",
+    log:
+        LOG_FP / "sga_abritamr_{sample}.log",
+    benchmark:
+        BENCHMARK_FP / "sga_abritamr_{sample}.tsv"
+    conda:
+        "envs/abritamr.yml"
+    shell:
+        """
+        abritamr run \\
+        --contigs {input.contigs} \\
+        --prefix {output.abritamr} \\
+        &> {log}
+        """