Update protocol

config/example_genomes.csv

@@ -9,5 +9,5 @@ Cryptococcus_neoformans.gff3, ,0,gzip,https://ftp.ncbi.nlm.nih.gov/genomes/all/G
 Tremella_mesenterica.faa, ,0,gzip,https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/271/645/GCF_000271645.1_Treme1/GCF_000271645.1_Treme1_protein.faa.gz,
 Wallemia_ichthyophaga.faa, ,0,gzip,https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/400/465/GCF_000400465.1_Wallemia_ichthyophaga_version_1.0/GCF_000400465.1_Wallemia_ichthyophaga_version_1.0_protein.faa.gz,
 Rhodotorula_graminis.faa, ,0,gzip,https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/001/329/695/GCF_001329695.1_Rhoba1_1/GCF_001329695.1_Rhoba1_1_protein.faa.gz,
-Saccharomyces_cerevisiae.faa, ,0,gzip,ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/146/045/GCF_000146045.2_R64/GCF_000146045.2_R64_protein.faa.gz,
+Saccharomyces_cerevisiae.faa, ,0,gzip,https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/146/045/GCF_000146045.2_R64/GCF_000146045.2_R64_protein.faa.gz,
 Rhizophagus_irregularis.faa, ,0,gzip,https://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/439/145/GCF_000439145.1_ASM43914v3/GCF_000439145.1_ASM43914v3_protein.faa.gz,

documentation/drac.md

@@ -0,0 +1,55 @@
+
+# How to install and run AMOEBAE on Digital Research Alliance of Canada (DRAC) clusters
+
+**Note**: This is mostly relevant for AMOEBAE users in Canada. However, this approach can be easily adapted to any high performance computing cluster.
+
+AMOEBAE depends on Conda packages, but installing Conda packages directly on DRAC isn't appropriate for the system. One way to circumvent that issue is to install dependencies in an Apptainer container, by performing the following steps.
+
+1. Log on to a DRAC cluster login node, clone the amoebae repository, and navigate into the amoebae directory. Upload your sequence data, as needed using an appropriate method (see [DRAC documentation on data transfers](https://docs.alliancecan.ca/wiki/Transferring_data)). Also, I recommend doing the subsequent steps in a screen or tmux session so that your work won't be interrupted if you get disconnected (see [DRAC documentation on prolonging terminal sessions](https://docs.alliancecan.ca/wiki/Prolonging_terminal_sessions)).
+
+2. Build the container using the `pixi.def` definition file provided in the AMOEBAE repository (see [DRAC Apptainer documentation](https://docs.alliancecan.ca/wiki/Apptainer)):
+    ```bash
+    module load apptainer
+    apptainer build --disable-cache pixi.sif pixi.def
+    ```
+    - This will take a few minutes.
+
+3. To run AMOEBAE workflow steps requiring internet access (`download_queries`
+   and `download_dbs`), enter a shell session within the
+   container on the login node (unless using the Cedar cluster specifically):
+    - Enter the shell session:
+        ```bash
+        APPTAINER_BIND=''
+        apptainer shell -C -B $PWD:/root --pwd /root pixi.sif
+        ```
+    - Now you are in an environment where all the AMOEBAE dependencies are
+      installed. You can run `snakemake` commands as described in the [workflow
+      protocol](./workflow_protocol.md), without the need to prefix with `pixi run `.
+    - Run the specific rules in the workflow (if/when needed as part of the workflow protocol):
+        ```bash
+        snakemake --cores 1 download_queries
+        snakemake --cores 1 download_dbs
+        ```
+    - Exit the shell session:
+        ```bash
+        exit
+        ```
+
+4. Otherwise, start an interactive session, so that your analysis will be run on a compute
+   node. For example, like this (see [DRAC documentation](https://docs.alliancecan.ca/wiki/Running_jobs#Interactive_jobs)):
+    ```bash
+    salloc --time=1:0:0 --mem-per-cpu=16G --ntasks=1 --account=def-leppard
+    ```
+
+5. Then, enter a shell session within the container:
+    ```bash
+    module load apptainer
+    APPTAINER_BIND=''
+    apptainer shell -C -B $PWD:/root --pwd /root pixi.sif
+    ```
+
+6. Now run the remaining workflow steps, as described in the workflow protocol. 
+    ```bash
+    snakemake get_ref_seqs 
+    ...
+    ```

documentation/workflow_protocol.md

@@ -4,21 +4,14 @@
 ## Requirements 
 
 The following setup procedure should work on most Linux or MacOS
-computers.  Please ensure that you have sufficient storage, as this protocol
+systems. Please ensure that you have sufficient storage, as this protocol
 will generate files totalling ~30GB or more in size.
 
-Most users seem to run the AMOEBAE workflow either on their personal computers
-or on servers without job schedulers (PBS, SLURM, etc.), so the instructions
-below will not be directly applicable to running AMOEBAE on a high-performance
-computing (HPC) cluster. If that is something you need to do, please refer to
-the documentation on the [Snakemake
-website](https://snakemake.readthedocs.io/en/stable/) and consult with your
-system administrator(s) as necessary. Otherwise, just follow the installation
-instructions below.
-
 
 ## Installation
 
+### Local servers and personal computers
+
 These instructions are for setting up and running AMOEBAE via the
 [SnakeMake](https://snakemake.readthedocs.io/en/stable/) command-line
 interface, which is well-documented and provides the flexibility to run AMOEBAE
@@ -75,6 +68,10 @@ installing dependencies on Apple Silicon MacOS systems.
     the Conda environment created using Pixi, and no longer use the
     `--use-conda` option. For example, `pixi run snakemake`.
 
+### High Performance Computing Clusters
+
+See the [instructions for using AMOEBAE on Digital Research Alliance of Canada clusters](./drac.md).
+
 ## Running the workflow
 
 With example (default) input files, this workflow should take between 30 and 60

pixi.def

@@ -0,0 +1,15 @@
+Bootstrap: docker
+From: ghcr.io/prefix-dev/pixi:latest
+
+%post
+    cd /root
+    cp pixi.toml pixi.lock /
+    cd /
+    pixi install --locked --environment default
+    pixi global install -c conda-forge -c bioconda blast muscle=3.8 exonerate
+
+%environment
+    export PIXI_HOME=/
+    export PIXI_CACHE_DIR=/.cache/rattler/cache
+    export RATTLER_AUTH_FILE=/.rattler/credentials.json
+    cd / && source <(pixi shell-hook --environment default)