Merge pull request #151 from nextstrain/open-data

Swap to open data for public forecasts

Author: joverlee521

.github/workflows/update-ncov-open-clade-counts.yaml

@@ -57,9 +57,7 @@ jobs:
 
   trigger_model_runs:
     needs: [open_clade_counts]
-    # Disable the run-model workflow due to low sequence counts
-    # <https://github.com/nextstrain/forecasts-ncov/issues/119>
-    if: false
+    if: ${{ !github.event.inputs.trial_name }}
     runs-on: ubuntu-latest
     steps:
       - run: gh workflow run run-models.yaml --repo nextstrain/forecasts-ncov -f data_provenance=open

config/config.yaml

@@ -1,6 +1,5 @@
 # See ingest/README.md for all available data
 data_provenances:
-  - gisaid
   - open
 variant_classifications:
   - nextstrain_clades
@@ -39,10 +38,10 @@ prepare_data:
     nextstrain_clades:
       global:
         included_days: 150
-        location_min_seq: 1000
+        location_min_seq: 200
         location_min_seq_days: 150
         excluded_locations: "defaults/global_excluded_locations.txt"
-        clade_min_seq: 500
+        clade_min_seq: 200
         clade_min_seq_days: 150
       usa:
         included_days: 150
@@ -53,12 +52,12 @@ prepare_data:
     pango_lineages:
       global:
         included_days: 150
-        location_min_seq: 1000
+        location_min_seq: 200
         location_min_seq_days: 150
         excluded_locations: "defaults/global_excluded_locations.txt"
         clade_min_seq: 1
         clade_min_seq_days: 150
-        collapse_threshold: 350
+        collapse_threshold: 200
 
 # Params for the model run scripts
 models:

viz/README.md

@@ -38,8 +38,8 @@ The command `npm run start:local` will both serve the website and provide access
 
 
 ```bash
-VITE_CLADES_PATH="gisaid/nextstrain_clades/global/mlr/YYYY-MM-DD_results.json" \
-  VITE_LINEAGES_PATH="gisaid/pango_lineages/global/mlr/YYYY-MM-DD_results.json" \
+VITE_CLADES_PATH="open/nextstrain_clades/global/mlr/YYYY-MM-DD_results.json" \
+  VITE_LINEAGES_PATH="open/pango_lineages/global/mlr/YYYY-MM-DD_results.json" \
   npm run start:local
 ```
 

viz/src/App.jsx

@@ -6,13 +6,13 @@ const mlrCladesConfig = {
     modelName: "mlr_clades",
     modelUrl: customAddress ?
       `${import.meta.env.VITE_DATA_HOST}/${import.meta.env.VITE_CLADES_PATH}` :
-      `https://nextstrain-data.s3.amazonaws.com/files/workflows/forecasts-ncov/gisaid/nextstrain_clades/global/mlr/latest_results.json`,
+      `https://nextstrain-data.s3.amazonaws.com/files/workflows/forecasts-ncov/open/nextstrain_clades/global/mlr/latest_results.json`,
 }
 const mlrLineagesConfig = {
     modelName: "mlr_lineages",
     modelUrl: customAddress ?
       `${import.meta.env.VITE_DATA_HOST}/${import.meta.env.VITE_LINEAGES_PATH}` :
-      `https://nextstrain-data.s3.amazonaws.com/files/workflows/forecasts-ncov/gisaid/pango_lineages/global/mlr/latest_results.json`,
+      `https://nextstrain-data.s3.amazonaws.com/files/workflows/forecasts-ncov/open/pango_lineages/global/mlr/latest_results.json`,
 }
 
 function App() {
@@ -32,17 +32,18 @@ function App() {
 
       <div className="warningContainer">
         <h2>
-          On Oct 1, 2025, we received an email from GISAID stating that they will no longer be updating the flat file
-          of EpiCoV data they've historically provisioned to Nextstrain since Feb 2020. Updates to these analyses are on
-          hold while we sort out alternative strategies to pull in new data from GISAID.
+          On Oct 1, 2025, we received an email from GISAID stating that they will no longer be updating the flat file of
+          EpiCoV data they've historically provisioned to Nextstrain since Feb 2020. Updates to GISAID-based analyses are
+          currently on hold. Please use this analysis based on open data instead, updates to which are continuing at their
+          regular weekly cadence.
         </h2>
       </div>
 
       <div id="mainPanelsContainer">
         <h2>Clade frequencies over time</h2>
         <p>
           Each line represents the estimated frequency of a particular clade through time. Equivalent Pango lineage is given
-          in parenthesis, eg clade 23A (lineage XBB.1.5). Only locations with more than 1000 sequences from samples collected
+          in parenthesis, eg clade 23A (lineage XBB.1.5). Only locations with more than 200 sequences from samples collected
           in the previous 150 days are included. Results last updated {mlrCladesData?.modelData?.get('updated') || 'loading'}.
         </p>
         <div id="cladeFrequenciesPanel" class="panelDisplay"> {/* surrounding div(s) used for static-images.js script */}
@@ -64,7 +65,7 @@ function App() {
         <h2>Lineage frequencies over time</h2>
         <p>
           Each line represents the estimated frequency of a particular Pango lineage through time. Lineages with fewer
-          than 350 observations are collapsed into parental lineage. Only locations with more than 1000 sequences from
+          than 200 observations are collapsed into parental lineage. Only locations with more than 200 sequences from
           samples collected in the previous 150 days are included. Results last updated
           {mlrLineagesData?.modelData?.get('updated') || 'loading'}.
         </p>