Merge pull request #291 from DendrouLab/fc_update_multivi

Fc update multivi

.github/workflows/integration01-ci.yml

@@ -37,12 +37,17 @@ jobs:
           activate-environment: pipeline_env
           environment-file: pipeline_env.yaml
 # important: this patch is only to test if multivi integration works
-# issues are not related to panpipes https://discourse.scverse.org/t/error-when-training-model-on-m3-max-mps/1896/2
+# pip install -e . #'.[multivipatch]' 
+# issues are not related to panpipes 
+# https://discourse.scverse.org/t/error-when-training-model-on-m3-max-mps/1896/2
 # https://discourse.scverse.org/t/macbook-m1-m2-mps-acceleration-with-scvi/2075/4
+
+# now disabling adversarial training for temp fix to ensure running latest multivi
+# https://github.com/scverse/scvi-tools/issues/2581
       - name: Install Panpipes
         shell: bash -el {0}
         run: |
-          pip install '.[multivipatch]' 
+          pip install -e . 
           conda list
 
       - name: Conda info
@@ -98,3 +103,7 @@ jobs:
       - name: File tree
         if: env.debug == 'true'
         run: tree teaseq
+
+      - name: check logs
+        if: always()
+        run: cat teaseq/integration/logs/4_multimodal_multivi.log

CHANGELOG.md

@@ -9,6 +9,8 @@
 
 - fixed calls to new imports matplotlib 
 - fixed scanpy/muon latest
+- changed default adversarial_training = False to ensure multivi runs with scvi-tools 1.1.3 https://github.com/scverse/scvi-tools/issues/2581
+- updated integration01 action to use updated scvi-tools version
 
 ### dependencies
 

docs/release_notes.md

@@ -9,6 +9,8 @@
 
 - fixed calls to new imports matplotlib 
 - fixed scanpy/muon latest
+- changed default adversarial_training = False to ensure multivi runs with scvi-tools 1.1.3 https://github.com/scverse/scvi-tools/issues/2581
+- updated integration01 action to use updated scvi-tools version
 
 ### dependencies
 

docs/yaml_docs/pipeline_integration_yml.md

@@ -350,7 +350,7 @@ For more information on `bbknn` check the [bbknn documentation](https://bbknn.re
         -  <span class="parameter">n_steps_kl_warmup</span> `String`, Default: None<br>
            Leave blank for the default integer
         -  <span class="parameter">n_epochs_kl_warmup</span> `Integer`, Default: 50<br>
-        -  <span class="parameter">adversarial_mixing</span> `Boolean`, Default: True<br>
+        -  <span class="parameter">adversarial_mixing</span> `Boolean`, Default: False<br>
   -  <span class="parameter">training_plan</span> `String`, Default: None<br>
 
 

panpipes/panpipes/pipeline_integration/pipeline.yml

@@ -190,7 +190,7 @@ multimodal:
       check_val_every_n_epoch :
       n_steps_kl_warmup :
       n_epochs_kl_warmup : 50
-      adversarial_mixing : True
+      adversarial_mixing : False
     training_plan :
 
   # Mofa arguments

panpipes/python_scripts/batch_correct_multivi.py

@@ -244,6 +244,14 @@
 
 mvi.view_anndata_setup()
 
+L.info("training args")
+print(multivi_training_args)
+
+
+L.info("training plan")
+print(multivi_training_plan)
+
+
 L.info("Running multiVI")
 mvi.train( **multivi_training_args, **multivi_training_plan)
 

pyproject.toml

@@ -71,11 +71,6 @@ spatial = [
     "tangram-sc"
 ]
 
-multivipatch = [
-    "scvi-tools<=0.20.3",
-    "requests"
-]
-
 refmap_old = [
     "scvi-tools",
     "pandas<2.0.0",

tests/integration_1/pipeline.yml

@@ -240,7 +240,6 @@ multimodal:
     # you can add any other param from the tutorials and they will
     # be parsed alongside the others
     # leave arguments blank for default
-    seed: 1492
     lowmem: True
     # Set lowmem to True will subset the atac to the top 25k HVF. 
     # This is to deal with concatenation of atac,rna on large datasets which at the moment is suboptimally required by scvitools.
@@ -260,14 +259,14 @@ multimodal:
       fully_paired : False 
     training_args:
       #(default: 500)
-      max_epochs : 500 
+      max_epochs : 20 
       #float (default: 0.0001)
       lr : 1.0e-05
-      #leave blanck for default str | int | bool | None (default: None)
+      #leave blank for default str | int | bool | None (default: None)
       use_gpu :
       # float (default: 0.9)
       train_size : 0.9 
-      # leave blanck for default, float | None (default: None)
+      # leave blank for default, float | None (default: None)
       validation_size : 
       # int (default: 128)
       batch_size : 128
@@ -279,15 +278,15 @@ multimodal:
       early_stopping : True 
       #bool (default: True)
       save_best : True
-       #leave blanck for default int | None (default: None)
+       #leave blank for default int | None (default: None)
       check_val_every_n_epoch :
-      #leave blanck for default int | None (default: None)
+      #leave blank for default int | None (default: None)
       n_steps_kl_warmup : 
        # int | None (default: 50)
       n_epochs_kl_warmup : 50
       #bool (default: True)
       adversarial_mixing : False 
-       #leave blanck for default dict | None (default: None)
+      #leave blank for default dict | None (default: None)
     training_plan :
   mofa:
     # this is a minimal set of parameters that will be expected