add ADMIXTURE example into factor analysis. remove CFA. redo PCA notebook accordingly.

.github/workflows/deploy-book.yml

@@ -23,7 +23,9 @@ jobs:
 
     - name: Build the book
       run: |
-        jupyter-book build . --config website/_config.yml --toc website/_toc.yml
+        cp website/_config.yml .
+        cp website/_toc.yml .
+        jupyter-book build .
 
     - name: GitHub Pages action
       uses: peaceiris/actions-gh-pages@v3.6.1

collider.ipynb

@@ -582,7 +582,19 @@
    "cell_type": "markdown",
    "id": "l9y51n6cmr",
    "metadata": {},
-   "source": "# Extended Reading\n\n**Aschard et al. (2015).** \"Adjusting for Heritable Covariates Can Bias Effect Estimates in Genome-Wide Association Studies.\" *American Journal of Human Genetics*, 96(3), 329-339. https://www.cell.com/ajhg/fulltext/S0002-9297(14)00527-8\n\nThis landmark paper demonstrates that adjusting for heritable covariates—traits with a genetic basis—can introduce significant bias into genome-wide association studies. The authors show that when genetic variants influence both an outcome of interest and the covariate being adjusted for, the effect estimates can be severely distorted, even creating false-positive associations. Using real data from large GWAS meta-analyses of anthropometric traits (waist-to-hip ratio and waist circumference adjusted for BMI), they found that approximately half of the reported associations could be explained by genetic effects on BMI rather than the primary outcome. The paper provides both theoretical framework and practical examples demonstrating that collider bias arises from shared causal factors between the outcome and covariate, and that standard adjustment methods cannot eliminate this bias when the underlying causal structure is unknown. This is essential reading for anyone conducting GWAS of correlated traits, as it highlights the importance of understanding causal relationships before deciding which variables to adjust for in your analysis."
+   "source": [
+    "# Extended Reading\n",
+    "\n",
+    "**Aschard et al. (2015).** \"Adjusting for Heritable Covariates Can Bias Effect Estimates in Genome-Wide Association Studies.\" *American Journal of Human Genetics*, 96(3), 329-339. https://www.cell.com/ajhg/fulltext/S0002-9297(14)00527-8\n",
+    "\n",
+    "This landmark paper demonstrates that adjusting for heritable covariates—traits with a genetic basis—can introduce significant bias into genome-wide association studies. The authors show that when genetic variants influence both an outcome of interest and the covariate being adjusted for, the effect estimates can be severely distorted, even creating false-positive associations. Using real data from large GWAS meta-analyses of anthropometric traits (waist-to-hip ratio and waist circumference adjusted for BMI), they found that approximately half of the reported associations could be explained by genetic effects on BMI rather than the primary outcome. The paper provides both theoretical framework and practical examples demonstrating that collider bias arises from shared causal factors between the outcome and covariate, and that standard adjustment methods cannot eliminate this bias when the underlying causal structure is unknown. This is essential reading for anyone conducting GWAS of correlated traits, as it highlights the importance of understanding causal relationships before deciding which variables to adjust for in your analysis."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "ecd50770",
+   "metadata": {},
+   "source": []
   }
  ],
  "metadata": {
@@ -603,4 +615,4 @@
  },
  "nbformat": 4,
  "nbformat_minor": 5
-}
+}

website/_toc.yml

@@ -23,7 +23,7 @@ parts:
   - caption: Correlation Between Individuals
     chapters:
       - file: genetic_relationship_matrix
-
+  
   - caption: Correlation Between Variables
     chapters:
       - file: factor_analysis

website/setting_up.md

@@ -21,7 +21,7 @@ The created template will have:
 
 
 ```bash
-jupyter-book build . --config website/_config.yml --toc website/_toc.yml
+jupyter-book build
 ```
 
 This will generate the HTML files in the `_build/html` directory.