Update ReadTheDocs

Author: serjisa

docs/source/tutorial.rst

@@ -4,21 +4,21 @@ Tutorial
 Step 1: Clonal nearest neighbours graph construction
 ****************************************************
 
-Firstly, we have to identify *k* nearest clonally labelled cells for each cell. It will create "bag of clones"
-(similar to "bag of words") that will be used for *clone2vec* training further.
+Firstly, we will create AnnData-object with clones.
 
 .. code-block:: python
 
     import scanpy as sc
     import sclitr as sl
 
-    sl.tl.clonal_nn(
+    clones = sl.pp.clones_adata(
         adata,
-        obs_name="clone",  # Column with clonal labels
-        use_rep="X_pca",   # Which dimred to use for graph construction
-        min_size=5,         # Minimal clone size
+        obs_name="clone", # Column with clonal labels
+        min_size=2,       # Minimal clone size
+        na_value="NA",    # Value for non-labelled cells
     )
 
+
 Minimal clone size parameter is used to exclude small clones from embedding construction.
 
 Step 2: clone2vec
@@ -29,24 +29,14 @@ clonally labelled cell.
 
 .. code-block:: python
 
-    clones = sl.pp.clones_adata(
+    sl.tl.clonal_nn(
         adata,
-        obs_name="clone",
-        min_size=5,
-        fill_obs="cell_type",  # Optional: composition column to fill clones layers
+        clones,
+        use_rep="X_pca", # Which dimred to use for graph construction
     )
 
-    # build clone graph in clones.obsp["gex_adjacency"] using cell-level embedding
-    # e.g., cosine similarity between clone centroids in `adata.obsm["X_pca"]` and kNN
-    # (see your pipeline; not shown here for brevity)
+    sl.tl.clone2vec(clones)
 
-    sl.tl.clone2vec(
-        clones,
-        z_dim=10,
-        obsp_key="gex_adjacency",  # graph between clones
-        obsm_key="clone2vec",
-        uns_key="clone2vec",
-    )
 
 After execution of this function we have AnnData-object :code:`clones` with clonal vector representation
 stored in :code:`clones.obsm["clone2vec"]`. Now we can work with it like with regular scRNA-Seq dataset.
@@ -62,29 +52,4 @@ Step 3: clone2vec analysis
 
 And after perform all other additional steps of analysis.
 
-Step 4: Identify predictors of clonal behaviour
-***********************************************
-
-In the simplest case, the model can be built to identify gene expression predictors of (a) position on a
-clonal embedding and (b) cell type composition of clones based on the expression in progenitor cells (if they exist).
-More broadly, we don't have to limit the prediction by the progenitor cells, and in this case the algorithm will
-identify general gene expression predictors of the distribution of the clone on an embedding.
-
-.. code-block:: python
-
-    shapdata_c2v = sl.tl.catboost(
-        adata,
-        obsm_key="X_umap",  # predict clone position; replace with your embedding
-        gs_key="gs",         # optional: use gs split info if available
-        model="regressor",
-        num_trees=1000,
-    )
-
-    shapdata_ct = sl.tl.associations(
-        adata,
-        response_key="proportions",    # or a specific layer/metric
-        response_field="obsm",         # depends on how you store targets
-        method="gam",                  # pearson/spearman/gam
-    )
-
 For a more detailed walkthrough see the Examples section.