Australian-Structural-Biology-Computing · jscgh · Mar 26, 2026 · Mar 26, 2026 · Mar 26, 2026 · Mar 26, 2026
diff --git a/.env.example b/.env.example
@@ -0,0 +1,28 @@
+# Copy this file to `template/.env` and adjust values for your site.
+# Runtime settings
+OOD_RESULTS_URL_BASE=/pun/sys/dashboard/files/fs
+PFOLD_SCRIPT_FILE=/path/to/proteinfold/bin/start-alphafold2.sh
+PFOLD_NATIVE_DEFAULT="-A your_project;-l select=1:ncpus=2:mem=4gb;-l walltime=48:00:00"
+PFOLD_NATIVE_GPU="-l select=1:ncpus=8:ngpus=1:mem=124gb;-l walltime=12:00:00"
+
+# Pipeline locations
+PFOLD_RUN_ENVIRONMENT=prod
+PFOLD_PROJECT_ROOT=/path/to/proteinfold
+PFOLD_DB_PATH=/path/to/proteinfold/dbs
+PFOLD_BRANCH=2.0.0
+PFOLD_REPOSITORY=nf-core/proteinfold
+PFOLD_NEXTFLOW_CONFIG=/path/to/proteinfold/nextflow.config
+
+# Runtime directories
+PFOLD_BASE_OUT_DIR=/path/to/output/proteinfold_output
+PFOLD_BASE_DEBUGDIR=/path/to/debug
+PFOLD_BASE_NXF_WORK=/path/to/work/proteinfold
+PFOLD_DEBUGGROUP=your_group
+
+# Container caches
+APPTAINER_CACHEDIR=/path/to/cache/apptainer
+SINGULARITY_CACHEDIR=/path/to/cache/singularity
+NXF_APPTAINER_CACHEDIR=/path/to/cache/apptainer
+NXF_SINGULARITY_CACHEDIR=/path/to/cache/singularity
+NXF_APPTAINER_LIBRARYDIR=/path/to/library/apptainer
+NXF_SINGULARITY_LIBRARYDIR=/path/to/library/singularity
diff --git a/.gitignore b/.gitignore
@@ -1 +1,2 @@
 .env
+template/.env
diff --git a/README.md b/README.md
@@ -33,51 +33,27 @@ To deploy ProteinFold on your own Open OnDemand instance:
    git clone https://github.com/Australian-Structural-Biology-Computing/ood-proteinfold.git /var/www/ood/apps/sys/kod-proteinfold
    ```
 
-2. **Edit files to match institutional settings**:
-   - Update `submit.yml.erb` with your facility-specific settings.
-   - Change the cluster in `form.yml.erb` to match your HPC.
-   - Create `kod_proteinfold-${RUN_ENVIRONMENT}.config` files for proteinfold-specific nextflow configurations. Or remove this line from `script.sh.erb` if not needed.
-   - Create a `template/.env` file to change any of these default parameters:
+2. **Configure for your site**:
+   - Copy `.env.example` to `template/.env`.
+   - Adjust the runtime environment variables in `template/.env` for paths, caches, output locations, and other job-time settings.
+   - Adjust the portal-side defaults at the top of the ERB files for cluster, queue, and other pre-submit settings.
+   - If needed, set `PFOLD_NEXTFLOW_CONFIG` to your institutional Nextflow config file instead of editing `script.sh.erb`.
 
 ```
-# Example .env for ood-proteinfold
-
-# Base output directory for all runs (job-specific OUT_DIR set at runtime)
-BASE_OUT_DIR=/srv/scratch/${USER}/proteinfold_output
-
-# Environment name (e.g., prod, dev)
-RUN_ENVIRONMENT=prod
-
-# Project root: location of Nextflow configs and samplesheet-utils environment
-PROJECT_ROOT=/srv/scratch/sbf-pipelines/proteinfold
-
-# Database path for structure prediction modules
-DB_PATH=/srv/scratch/sbf-pipelines/proteinfold/dbs
-
-# Git branch to use for proteinfold
-BRANCH=master
-
-# Repository path for using alternative versions of proteinfold
-REPOSITORY=Australian-Structural-Biology-Computing/proteinfold
-
-# Debug group for permissions on debug files
-DEBUGGROUP=sbf-pipelines
-
-# Base debug directory (job-specific DEBUGDIR set at runtime)
-BASE_DEBUGDIR=/srv/scratch/sbf/debug
-
-# Nextflow work directory base (job-specific NXF_WORK set at runtime)
-BASE_NXF_WORK=/srv/scratch/${USER}/.proteinfold/work
-
-# Apptainer/Singularity blob cache directories (set here or via your institutional Nextflow config, e.g. https://github.com/Australian-Structural-Biology-Computing/unsw_katana)
-# APPTAINER_CACHEDIR=/srv/scratch/${USER}/.images
-# SINGULARITY_CACHEDIR=/srv/scratch/${USER}/.images
-
-# Nextflow Apptainer/Singularity image cache/library directories (set here or via institutional Nextflow config, e.g. https://github.com/nf-core/configs/blob/master/conf/pipeline/proteinfold/unsw_katana.config)
-# NXF_APPTAINER_CACHEDIR=/srv/scratch/${USER}/.images
-# NXF_SINGULARITY_CACHEDIR=/srv/scratch/${USER}/.images
-# NXF_APPTAINER_LIBRARYDIR=/srv/scratch/sbf-pipelines/proteinfold/singularity
-# NXF_SINGULARITY_LIBRARYDIR=/srv/scratch/sbf-pipelines/proteinfold/singularity
+# Example template/.env for ood-proteinfold
+
+PFOLD_SCRIPT_FILE=/path/to/proteinfold/bin/start-alphafold2.sh
+PFOLD_NATIVE_DEFAULT=-A my_project;-l select=1:ncpus=2:mem=4gb;-l walltime=48:00:00
+PFOLD_NATIVE_GPU=-l select=1:ncpus=8:ngpus=1:mem=124gb;-l walltime=12:00:00
+PFOLD_RUN_ENVIRONMENT=prod
+PFOLD_PROJECT_ROOT=/apps/proteinfold
+PFOLD_DB_PATH=/data/proteinfold/dbs
+PFOLD_BRANCH=master
+PFOLD_REPOSITORY=Australian-Structural-Biology-Computing/proteinfold
+PFOLD_NEXTFLOW_CONFIG=/apps/proteinfold/site.config
+PFOLD_BASE_OUT_DIR=/scratch/${USER}/proteinfold_output
+PFOLD_BASE_NXF_WORK=/scratch/${USER}/.proteinfold/work
+OOD_RESULTS_URL_BASE=/pun/sys/dashboard/files/fs
 ```
 
 ### Python Environment for samplesheet-utils
@@ -91,7 +67,13 @@ source ${PROJECT_ROOT}/${RUN_ENVIRONMENT}/venv/bin/activate
 pip install samplesheet-utils==1.1.2
 ```
 
-Update the paths in `.env` if your environment is elsewhere.
+Prefer changing `template/.env` over editing the app templates directly so upgrades stay easy to rebase across institutions.
+
+Portal rendering and job runtime are configured differently:
+- portal-side defaults live at the top of `form.yml.erb`, `submit.yml.erb`, `info.html.erb`, `completed.html.erb`, and `view.html.erb`
+- runtime config lives in `template/.env`, which is intentionally gitignored
+
+For CI/CD, prefer generating `template/.env` during deployment from site-managed config or secrets rather than committing a real site config file to the repository.
 
 ## Citation
 

diff --git a/completed.html.erb b/completed.html.erb
@@ -1,5 +1,8 @@
 <% if !connect.user.blank? %>
 <%
+portal_default_base_out_dir = "/srv/scratch/#{connect.user}/proteinfold_output"
+portal_default_results_url_base = "/pun/sys/dashboard/files/fs"
+
 def file_url(base, path)
   encoded_path = path.to_s.split("/").reject(&:empty?).map { |segment| ERB::Util.url_encode(segment) }.join("/")
   "#{base}/#{encoded_path}"
@@ -69,8 +72,8 @@ def sanitize_live_log(text)
   compact.join("\n").strip
 end
 
-base_out_dir = (connect.respond_to?(:base_out_dir) && !connect.base_out_dir.blank?) ? connect.base_out_dir : "/srv/scratch/#{connect.user}/proteinfold_output"
-results_url_base = (connect.respond_to?(:results_url_base) && !connect.results_url_base.blank?) ? connect.results_url_base : "/pun/sys/dashboard/files/fs"
+base_out_dir = (connect.respond_to?(:base_out_dir) && !connect.base_out_dir.blank?) ? connect.base_out_dir : portal_default_base_out_dir
+results_url_base = (connect.respond_to?(:results_url_base) && !connect.results_url_base.blank?) ? connect.results_url_base : portal_default_results_url_base
 session_output_dir = (connect.respond_to?(:session_output_dir) && !connect.session_output_dir.blank?) ? connect.session_output_dir : nil
 log_dir = session_output_dir || File.join(base_out_dir, connect.run_dir)
 live_log_path = File.join(log_dir, "live_log.log")
@@ -82,7 +85,7 @@ summary_log_label = File.file?(output_log_path) ? "output.log" : "live_log.log"
 log_text = sanitize_live_log(tail_text(live_log_path))
 %>
 <div class="ood-completed-log-card">
-<div><b>Log</b>: <a target="_blank" rel="noopener" href="https://kod.restech.unsw.edu.au<%= summary_log_url %>"><%= summary_log_label %></a></div>
+<div><b>Log</b>: <a target="_blank" rel="noopener" href="<%= summary_log_url %>"><%= summary_log_label %></a></div>
 <% if log_text && !log_text.empty? %>
 <pre class="ood-completed-log" style="margin-top: 0.75rem !important; height: 7rem !important; max-height: 7rem !important; overflow: auto !important; padding: 1rem !important; background: #0f172a !important; color: #f8fafc !important; border: 1px solid #334155 !important; border-radius: 0.5rem !important; white-space: pre-wrap !important; font: 0.9rem/1.45 ui-monospace, SFMono-Regular, Menlo, Monaco, Consolas, Liberation Mono, monospace !important;"><%= ERB::Util.html_escape(log_text) %></pre>
 <script>

diff --git a/form.yml.erb b/form.yml.erb
@@ -1,3 +1,7 @@
+<%
+portal_default_cluster = "katana"
+%>
+
 form:
     - data_warning
     - samplesheet
@@ -13,7 +17,7 @@ form:
     - citation_request
     - katana_citation
 
-cluster: "katana"
+cluster: "<%= portal_default_cluster %>"
 
 attributes:
     data_warning:
@@ -32,7 +36,7 @@ attributes:
         class: "d-none"
         label: "Input Options"
         help: |
-            • <b>Directory:</b> <code>/srv/scratch/z3141592/my_experiment/</code>
+            • <b>Directory:</b> <code>/path/to/my_experiment/</code>
             <br>• <b>Single file:</b> <code>protein.fasta</code>, <code>protein.yaml</code>, or <code>samplesheet.csv</code>
             <br>• <b>Sequence:</b> <code>NLYIQWLKDGGPSSGRPPPS</code> or <code>NLYIQWLKDGGPSSGRPPPS:MKTVRQERLKSDHELFRTHT</code>
             <br>📖 <a href="https://unsw.sharepoint.com/sites/StructuralBiologyFacility-UserDocumentation/SitePages/ProteinFold-User-Guide.aspx?ga=1#preparing-your-inputs">User Guide</a> for advanced options.

diff --git a/info.html.erb b/info.html.erb
@@ -1,12 +1,15 @@
 <% if !connect.user.blank? %>
 <%
+portal_default_base_out_dir = "/srv/scratch/#{connect.user}/proteinfold_output"
+portal_default_results_url_base = "/pun/sys/dashboard/files/fs"
+
 def file_url(base, path)
   encoded_path = path.to_s.split("/").reject(&:empty?).map { |segment| ERB::Util.url_encode(segment) }.join("/")
   "#{base}/#{encoded_path}"
 end
 
-base_out_dir = (connect.respond_to?(:base_out_dir) && !connect.base_out_dir.blank?) ? connect.base_out_dir : "/srv/scratch/#{connect.user}/proteinfold_output"
-results_url_base = (connect.respond_to?(:results_url_base) && !connect.results_url_base.blank?) ? connect.results_url_base : "/pun/sys/dashboard/files/fs"
+base_out_dir = (connect.respond_to?(:base_out_dir) && !connect.base_out_dir.blank?) ? connect.base_out_dir : portal_default_base_out_dir
+results_url_base = (connect.respond_to?(:results_url_base) && !connect.results_url_base.blank?) ? connect.results_url_base : portal_default_results_url_base
 %>
-<b>Results</b>: <a target="_blank" rel="noopener" href="https://kod.restech.unsw.edu.au<%= file_url(results_url_base, File.join(base_out_dir, connect.run_dir)) %>"><%= connect.run_dir %></a>
+<b>Results</b>: <a target="_blank" rel="noopener" href="<%= file_url(results_url_base, File.join(base_out_dir, connect.run_dir)) %>"><%= connect.run_dir %></a>
 <% end %>
diff --git a/submit.yml.erb b/submit.yml.erb
@@ -1,7 +1,34 @@
+<%
+portal_default_cluster = "katana"
+portal_default_queue = "submission"
+portal_default_script_path = "/srv/scratch/sbf-pipelines/proteinfold/bin/start-alphafold2.sh"
+portal_default_native = "-A kod_proteinfold;-l select=1:ncpus=2:mem=4gb;-l walltime=48:00:00"
+portal_default_gpu_native = "-l select=1:mwacgpu200=gpu-mwac:ncpus=8:ngpus=1:mem=124gb;-l walltime=12:00:00"
+portal_default_email_domain = "@ad.unsw.edu.au"
+
+cluster_name = portal_default_cluster
+queue_name = portal_default_queue
+script_path = portal_default_script_path
+default_native = portal_default_native
+gpu_native = portal_default_gpu_native
+
+def parse_native(val)
+    val.split(/\s*;\s*/).reject(&:empty?)
+end
+
+native_default_list = parse_native(default_native)
+native_gpu_list = parse_native(gpu_native)
+
+email_default = "#{ENV['USER']}#{portal_default_email_domain}"
+email_selected = (defined?(email) && email && !email.strip.empty?) ? email : email_default
+%>
+
+cluster: "<%= cluster_name %>"
+
 batch_connect:
     template: "basic"
-    queue_name: "submission"
-    script_file: "/srv/scratch/sbf-pipelines/proteinfold/bin/start-alphafold2.sh"
+    queue_name: "<%= queue_name %>"
+    script_file: "<%= script_path %>"
     conn_params:
         - base_out_dir
         - results_url_base
@@ -13,15 +40,16 @@ script:
     native:
     # ColabFold local DBs are only available locally on certain nodes. This is an example method for catching local ColabFold runs
      <% if (af_method == "colabfold" || af_method == "boltz") && msa_server == "local" %>
-         - "-l select=1:mwacgpu200=gpu-mwac:ncpus=8:ngpus=1:mem=124gb"
-         - "-l walltime=12:00:00"
+        <% native_gpu_list.each do |entry| %>
+         - "<%= entry %>"
+        <% end %>
          - "-N <%= (run_name || 'kod_proteinfold').gsub(' ', '_') %>"
      <% else %>
-        - "-A kod_proteinfold"
-        - "-l select=1:ncpus=2:mem=4gb"
-        - "-l walltime=48:00:00"
+        <% native_default_list.each do |entry| %>
+        - "<%= entry %>"
+        <% end %>
         - "-N <%= (run_name || 'kod_proteinfold').gsub(' ', '_') %>"
      <% end %>
     email:
-        - "<%= ENV['USER'] %>@ad.unsw.edu.au"
+        - "<%= email_selected %>"
     email_on_terminated: "<%= email_on_terminated %>"
diff --git a/template/after.sh.erb b/template/after.sh.erb
@@ -1,9 +1,25 @@
-source .env
+ENV_FILE=".env"
+if [ -f "$ENV_FILE" ]; then
+    set -a
+    . "$ENV_FILE"
+    set +a
+fi
+
+require_var() {
+    local name="$1"
+    if [ -z "${!name:-}" ]; then
+        echo "ERROR: Required config variable '$name' is not set. Define it in template/.env or the environment." >&2
+        exit 1
+    fi
+}
+
+require_var PFOLD_BASE_OUT_DIR
+require_var OOD_RESULTS_URL_BASE
 
 RUN_DIR=$(echo "<%= context.run_name %>" | sed "s#[^A-Za-z0-9]#_#g")
-BASE_OUT_DIR="${BASE_OUT_DIR:-/srv/scratch/${USER}/proteinfold_output}"
+BASE_OUT_DIR="${BASE_OUT_DIR:-${PFOLD_BASE_OUT_DIR}}"
 OUT_DIR="${BASE_OUT_DIR}/${RUN_DIR}"
-RESULTS_URL_BASE="${RESULTS_URL_BASE:-/pun/sys/dashboard/files/fs}"
+RESULTS_URL_BASE="${RESULTS_URL_BASE:-${OOD_RESULTS_URL_BASE}}"
 
 ln -s "$OUT_DIR" 2>/dev/null || true
 

diff --git a/template/script.sh.erb b/template/script.sh.erb
@@ -1,13 +1,44 @@
 #!/bin/bash
 
-source .env
-# Set defaults to prod if .env is not found
-RUN_ENVIRONMENT=${RUN_ENVIRONMENT:-prod}
-PROJECT_ROOT=${PROJECT_ROOT:-/srv/scratch/sbf-pipelines/proteinfold}
-DB_PATH=${DB_PATH:-/srv/scratch/sbf-pipelines/proteinfold/dbs}
-BRANCH=${BRANCH:-master}
-REPOSITORY=${REPOSITORY:-Australian-Structural-Biology-Computing/proteinfold}
-DEBUGGROUP=${DEBUGGROUP:-sbf}
+ENV_FILE=".env"
+if [ -f "$ENV_FILE" ]; then
+    set -a
+    . "$ENV_FILE"
+    set +a
+fi
+
+require_var() {
+    local name="$1"
+    if [ -z "${!name:-}" ]; then
+        echo "ERROR: Required config variable '$name' is not set. Define it in template/.env or the environment." >&2
+        exit 1
+    fi
+}
+
+require_var PFOLD_RUN_ENVIRONMENT
+require_var PFOLD_PROJECT_ROOT
+require_var PFOLD_DB_PATH
+require_var PFOLD_BRANCH
+require_var PFOLD_REPOSITORY
+require_var PFOLD_DEBUGGROUP
+require_var PFOLD_NEXTFLOW_CONFIG
+require_var PFOLD_BASE_DEBUGDIR
+require_var PFOLD_BASE_NXF_WORK
+require_var PFOLD_BASE_OUT_DIR
+require_var APPTAINER_CACHEDIR
+require_var SINGULARITY_CACHEDIR
+require_var NXF_APPTAINER_CACHEDIR
+require_var NXF_SINGULARITY_CACHEDIR
+require_var NXF_APPTAINER_LIBRARYDIR
+require_var NXF_SINGULARITY_LIBRARYDIR
+
+RUN_ENVIRONMENT=${RUN_ENVIRONMENT:-${PFOLD_RUN_ENVIRONMENT}}
+PROJECT_ROOT=${PROJECT_ROOT:-${PFOLD_PROJECT_ROOT}}
+DB_PATH=${DB_PATH:-${PFOLD_DB_PATH}}
+BRANCH=${BRANCH:-${PFOLD_BRANCH}}
+REPOSITORY=${REPOSITORY:-${PFOLD_REPOSITORY}}
+DEBUGGROUP=${DEBUGGROUP:-${PFOLD_DEBUGGROUP}}
+NEXTFLOW_CONFIG=${NEXTFLOW_CONFIG:-${PFOLD_NEXTFLOW_CONFIG}}
 
 set -xo pipefail
 set -eE
@@ -16,13 +47,13 @@ set -eE
 NOW=$(date +%Y%m%d_%H%M%S)
 RUNNAME="${NOW}_${USER}"
 
-BASE_DEBUGDIR=${BASE_DEBUGDIR:-"/srv/scratch/sbf/debug"}
+BASE_DEBUGDIR=${BASE_DEBUGDIR:-${PFOLD_BASE_DEBUGDIR}}
 DEBUGDIR="${BASE_DEBUGDIR}/${RUNNAME}"
 
-BASE_NXF_WORK=${BASE_NXF_WORK:-"/srv/scratch/${USER}/.proteinfold/work"}
+BASE_NXF_WORK=${BASE_NXF_WORK:-${PFOLD_BASE_NXF_WORK}}
 export NXF_WORK="${BASE_NXF_WORK}/${RUNNAME}"
 
-BASE_OUT_DIR=${BASE_OUT_DIR:-"/srv/scratch/${USER}/proteinfold_output"}
+BASE_OUT_DIR=${BASE_OUT_DIR:-${PFOLD_BASE_OUT_DIR}}
 RUN_DIR=$(echo "<%= context.run_name %>" | sed "s#[^A-Za-z0-9]#_#g")
 OUT_DIR="${BASE_OUT_DIR}/${RUN_DIR}"
 mkdir -p "${OUT_DIR}"
@@ -47,14 +78,14 @@ VERSION_SAMPLESHEET_UTILS="1.3.4"
 source "${PROJECT_ROOT}/${RUN_ENVIRONMENT}/venv/bin/activate"
 
 # Native environment cacheDir used for caching blobs during image pulls
-export APPTAINER_CACHEDIR="${APPTAINER_CACHEDIR:-/srv/scratch/${USER}/.images}"
-export SINGULARITY_CACHEDIR="${SINGULARITY_CACHEDIR:-/srv/scratch/${USER}/.images}"
+export APPTAINER_CACHEDIR="${APPTAINER_CACHEDIR}"
+export SINGULARITY_CACHEDIR="${SINGULARITY_CACHEDIR}"
 
 # Nextflow variables, checks libraryDir and if image is not found downloads to cache
-export NXF_APPTAINER_CACHEDIR="${NXF_APPTAINER_CACHEDIR:-/srv/scratch/${USER}/.images}"
-export NXF_SINGULARITY_CACHEDIR="${NXF_SINGULARITY_CACHEDIR:-/srv/scratch/${USER}/.images}"
-export NXF_APPTAINER_LIBRARYDIR="${NXF_APPTAINER_LIBRARYDIR:-/srv/scratch/sbf-pipelines/proteinfold/singularity}"
-export NXF_SINGULARITY_LIBRARYDIR="${NXF_SINGULARITY_LIBRARYDIR:-/srv/scratch/sbf-pipelines/proteinfold/singularity}"
+export NXF_APPTAINER_CACHEDIR="${NXF_APPTAINER_CACHEDIR}"
+export NXF_SINGULARITY_CACHEDIR="${NXF_SINGULARITY_CACHEDIR}"
+export NXF_APPTAINER_LIBRARYDIR="${NXF_APPTAINER_LIBRARYDIR}"
+export NXF_SINGULARITY_LIBRARYDIR="${NXF_SINGULARITY_LIBRARYDIR}"
 
 export PROTEINFOLD_VERSION="2.0.0dev"
 
@@ -215,15 +246,15 @@ fi
 
 if [ "${MODE}" == "FASTA_INPUT" ];
 then
-    mkdir fasta
+    mkdir -p fasta
     cp "${user_input}" ./fasta/
     create-samplesheet --directory "./fasta" --suffix "${prot_mode}" ${OUTPUT_FORMAT}
     SAMPLESHEET="$(pwd)/samplesheet.csv"
 fi
 
 if [ "${MODE}" == "DIRECTORY" ];
 then
-    mkdir fasta
+    mkdir -p fasta
     working_dir=$(pwd)
     cd "${user_input}"
     for filename in *
@@ -289,7 +320,7 @@ fi
 [[ "<%= context.af_method %>" == "rosettafold_all_atom" ]] && PROT_MODE="rosettafold_all_atom" && ARGS="${ARGS} --rosettafold_all_atom_db ${DB_PATH}"
 [[ "<%= context.af_method %>" == "helixfold3"  ]] && PROT_MODE="helixfold3" && ARGS="${ARGS} --helixfold3_db ${DB_PATH}"
 
-nextflow -c "${PROJECT_ROOT}/kod_proteinfold-${RUN_ENVIRONMENT}.config" run ${REPOSITORY} -r ${BRANCH} -latest \
+nextflow -c "${NEXTFLOW_CONFIG}" run ${REPOSITORY} -r ${BRANCH} -latest \
     --input "${SAMPLESHEET}" \
     --outdir "${OUT_DIR}" \
     --mode "${PROT_MODE}" \