spellcheck functional tutorials

README.md

@@ -10,7 +10,7 @@ Python package for inference and analysis of mutation trees.
 
 PYggdrasil implements the [Single Cell Inference of Tumor Evolution (SCITE)](https://github.com/cbg-ethz/SCITE) algortihm by [Kuipers J et al. (2015)](https://pubmed.ncbi.nlm.nih.gov/29030470/).
 
-It was designed to quantify the MCMC exploration of tumour progression tree spaces, in particular to investigate: Initialisation Strategies, Convergence Diagnostics & Multi-modalities of SCITE.
+It was designed to quantify the MCMC exploration of tumour progression tree spaces, in particular to investigate: Initialisation Strategies, Convergence Diagnostics & Multi-modalities.
 
 ## Usage
 
@@ -20,7 +20,7 @@ import pyggdrasil as yg
 
 
 ## Contributing
-See [Contributing Guidlines](https://cbg-ethz.github.io/PYggdrasil/contributing/).
+See [Contributing Guidelines](https://cbg-ethz.github.io/PYggdrasil/contributing/).
 ### Setting up the repository
 
 To build package and maintain dependencies we use [Poetry](https://python-poetry.org/).
@@ -73,5 +73,5 @@ We recommend submitting small pull requests and starting with drafts outlining p
 * Experimental workflows are in ``workflows/``, with a description of how to set up the environment in ``workflows/README.md``
 
 ## Origin & Authorship
-This pakage originates from [Gordon J Köhn](https://github.com/gordonkoehn)'s MSc Thesis: _[Quantifying MCMC Exploration of Tumour Progression Tree Spaces](TODO(Gordon):add in link)_ in 2023 at ETH Zürich.
-The project was supervised by [Paweł Czyż](https://pawel-czyz.github.io/) and Prof. Dr. Niko Beerenwinkel of the Combutational Biology Group at the [Department of Biosystems Science and Engineering](https://www.bsse.ethz.ch/).
+This package originates from [Gordon J Köhn](https://github.com/gordonkoehn)'s MSc Thesis: _[Quantifying MCMC Exploration of Tumour Progression Tree Spaces](TODO(Gordon): add in link)_ in 2023 at ETH Zürich.
+[Paweł Czyż](https://pawel-czyz.github.io/) and Prof. Dr Niko Beerenwinkel supervised this project as part of the Computational Biology Group at the [Department of Biosystems Science and Engineering](https://www.bsse.ethz.ch/).

docs/tutorial/analyzeMCMC.md

@@ -28,13 +28,12 @@ import matplotlib.pyplot as plt
 
 ## Run MCMC
 
-Below we run 4 Markov Chains, for 100 iterations each, with different
+Below we run 4 Markov Chains, for 200 iterations each, with different
 initial trees.
 
 ### Generate a ground-truth mutation history and a noisy single-cell mutation profile
 
-The below cell generates a random tree with 4 mutations, plus root. For
-debugging we may use the *print_topo* to plot its topology.
+The below cell generates a random tree with 4 mutations, plus root. 
 
 <details>
 <summary>Code</summary>
@@ -104,9 +103,9 @@ mut_mat = jnp.array(data['noisy_mutation_mat'])
 
 ## Run the Markov Monte Carlo Chain
 
-The below cell runs a 4 differnt MCMC chain. We initialize ti with the
+The below cell runs a 4 differnt MCMC chain. We initialize it with the
 initial tree from before. We configure the move probabilities and error
-rates and run the MCMC chain for 100 iterations. The sampels are saved
+rates and run the MCMC chain for 200 iterations. The sampels are saved
 to disk and loaded back into memory as chains may be very long.
 
 ``` python

docs/tutorial/analyzeMCMC.qmd

@@ -26,10 +26,10 @@ import matplotlib.pyplot as plt
 ```
 
 ## Run MCMC
-Below we run 4 Markov Chains, for 100 iterations each, with different initial trees.
+Below we run 4 Markov Chains, for 200 iterations each, with different initial trees.
 
 ### Generate a ground-truth mutation history and a noisy single-cell mutation profile
-The below cell generates a random tree with 4 mutations, plus root. For debugging we may use the _print_topo_ to plot its topology.
+The below cell generates a random tree with 4 mutations, plus root. 
 ```{python}
 #| code-fold: true 
 # make true tree
@@ -83,8 +83,8 @@ mut_mat = jnp.array(data['noisy_mutation_mat'])
 ```
 
 ##  Run the Markov Monte Carlo Chain
-The below cell runs a 4 differnt MCMC chain. We initialize ti with the initial tree from before.
-We configure the move probabilities and error rates and run the MCMC chain for 100 iterations.
+The below cell runs a 4 differnt MCMC chain. We initialize it with the initial tree from before.
+We configure the move probabilities and error rates and run the MCMC chain for 200 iterations.
 The sampels are saved to disk and loaded back into memory as chains may be very long.
 
 ```{python}

docs/tutorial/index.md

@@ -1,21 +1,21 @@
 # Tutorials
 
-We'd love to see you built upon PYggdrasil. Below we provide some tutorials to help you get started.
+We'd love to see you built upon PYggdrasil. Below, we provide some tutorials to help you get started.
 
 ## Functional Usage
-The below tutorials illustrate how to use PYggdrasil functions for specific tasks.
+The tutorials below illustrate how to use PYggdrasil functions for specific tasks.
 
-- [Single MCMC Run](singleMCMC.md) demonstrated how to run a single MCMC chain.
-- [Tree Similarities and Visualization](similarities.md) demonstrated how to compute distances between two trees and visualize trees.
-- [Analyzing MCMC Runs](analyzeMCMC.md) demonstrated how to compute distances between of MCMC chains and diagnose convergence issues.
+- [Single MCMC Run](singleMCMC.md) demonstrates how to run a single MCMC chain.
+- [Tree Similarities and Visualization](similarities.md) demonstrates how to compute distances between two trees and visualize trees.
+- [Analyzing MCMC Runs](analyzeMCMC.md) demonstrates how to compute distances between MCMC chains and diagnose convergence issues.
 
 
 ## Experimental Workflows
-For a sustaibalbe and reproducible workflow use _snakemake_ to run PYggdrasil.
+We use _snakemake_ to create a sustainable and reproducible workflow for extensive experiments with PYggdrasil.
 See [Workflows](../workflows/index.md) for more details.
 
-The below tutorials illustrate _snakemake_ workflows ilustrated that allow to run fully reproducible experiments using PYggdrasil.
-From tree and data generation to convergence diagostics a pipline exists.
+The tutorials below illustrate _snakemake_ workflows that allow running such fully reproducible experiments using PYggdrasil.
+From tree and data generation to convergence diagnostics, a pipeline exists.
 
-- [Basic Workflow](basicWorkflows.md) demonstrated how to run a basic workflow for a particular output file.
-- [Advanced Workflow](advanced_workflow.md) demonstrated how to run a more advanced workflow. 
+- [Basic Workflow](basicWorkflows.md) demonstrates a basic workflow for a particular output file.
+- [Advanced Workflow](advanced_workflow.md) demonstrates how to run a more advanced workflow. 

docs/tutorial/similarities.md

@@ -108,7 +108,7 @@ yg.visualize.plot_tree_no_print(deep_tree, save_name, save_dir)
 methods.
 
 1.  MCMC tree generation - takes a tree and evolves it by a fixed number
-    random moves implemnted with SCITE.
+    of random moves implemnted with SCITE.
 2.  HUNTRESS inference - takes a cell-mutation profile and infers a tree
     with HUNTRESS.
 

docs/tutorial/similarities.qmd

@@ -85,7 +85,7 @@ yg.visualize.plot_tree_no_print(deep_tree, save_name, save_dir)
 
 **Note:** PYggdrasil inplements two more advanced tree generation methods. 
 
-1. MCMC tree generation - takes a tree and evolves it by a fixed number random moves implemnted with SCITE.
+1. MCMC tree generation - takes a tree and evolves it by a fixed number of random moves implemnted with SCITE.
 2. HUNTRESS inference - takes a cell-mutation profile and infers a tree with HUNTRESS.
 
 

docs/tutorial/singleMCMC.md

@@ -3,7 +3,7 @@
 This tutorial shows how to run a single MCMC chain of SCITE using
 PYggdrasil.
 
-- We will generate our own ground-truth mutation histroy and generate a
+- We will generate our own ground-truth mutation history and generate a
   noisy single-cell mutation profile from it.
 - We will then run a single MCMC chain to infer the mutation history
   from the noisy single-cell mutation profile.
@@ -99,10 +99,10 @@ print(mut_mat)
 
 ## 4) Run the Markov Monte Carlo Chain
 
-The below cell runs a single MCMC chain. We initialize ti with the
-initial tree from before. We configure the move probabilities and error
-rates and run the MCMC chain for 100 iterations. The sampels are saved
-to disk and loaded back into memory as chains may be very long.
+The below cell runs a single MCMC chain. We initialize it with
+the initial tree from before. We configure the move probabilities
+and error rates and run the MCMC chain for 100 iterations. 
+The samples are saved to disk and loaded back into memory as chains may be very long.
 
 ``` python
 ## Run MCMC
@@ -145,10 +145,10 @@ mcmc_data = yg.serialize.read_mcmc_samples(save_dir / f"{save_name}.json")
 
 ## 5) Visualize the results
 
-In the following we would like to plot the evolution of the MCMC chain
-and the trees that were sampled. First we convert the data from the
-serialized format to a *pureMCMCdata* format. This is a simple dataclass
-that contains the trees and the log probabilities of the trees.
+In the following, we would like to plot the evolution of the MCMC chain
+and the trees that were sampled. First, we convert the data from the serialized
+format to a pureMCMCdata format. This is a simple data class that 
+contains the trees and the log probabilities of the trees.
 
 ``` python
 # unpack the data - reads in the serialized trees to Tree objects
@@ -198,11 +198,11 @@ yg.compare_trees(last_tree, true_tree)
 
     True
 
-Not note that the last tree does not need to be a good tree. SCITE is
-just likely to spend more iterations exploring more likely trees. Here
-the last tree just turns out to be a tree with the highest
-log-probability.
+Now note that the last tree does not need to be a good tree.
+SCITE is just likely to spend more iterations exploring more
+likely trees. Here, the last tree just turns out to be a
+tree with the highest log-probability.
 
-To acutally retrive a mutation tree from the posterior one would have to
-make a point estimate Maximum A Posteriori (MAP) tree, i.e. sampled the
-most times. See SCITE paper for details.
+To acutally retrive a mutation tree from the posterior one 
+would have to make a point estimate Maximum A Posteriori (MAP) tree,
+ i.e. sampled the most times.  See SCITE paper for details.

docs/tutorial/singleMCMC.qmd

@@ -6,7 +6,7 @@ jupyter: python3
 
 This tutorial shows how to run a single MCMC chain of SCITE using PYggdrasil. 
 
-* We will generate our own ground-truth mutation histroy and generate a noisy single-cell mutation profile from it.
+* We will generate our own ground-truth mutation history and generate a noisy single-cell mutation profile from it.
 * We will then run a single MCMC chain to infer the mutation history from the noisy single-cell mutation profile.
 * Visualize the results. The trees and the evolution of the MCMC.
 
@@ -77,9 +77,7 @@ print(mut_mat)
 ```
 
 ## 4) Run the Markov Monte Carlo Chain
-The below cell runs a single MCMC chain. We initialize ti with the initial tree from before.
-We configure the move probabilities and error rates and run the MCMC chain for 100 iterations.
-The sampels are saved to disk and loaded back into memory as chains may be very long.
+The below cell runs a single MCMC chain. We initialize it with the initial tree from before. We configure the move probabilities and error rates and run the MCMC chain for 100 iterations. The samples are saved to disk and loaded back into memory as chains may be very long.
 
 ```{python}
 #| warning: false
@@ -122,8 +120,10 @@ mcmc_data = yg.serialize.read_mcmc_samples(save_dir / f"{save_name}.json")
 ```
 
 ## 5) Visualize the results
-In the following we would like to plot the evolution of the MCMC chain and the trees that were sampled.
-First we convert the data from the serialized format to a _pureMCMCdata_ format. This is a simple dataclass that contains the trees and the log probabilities of the trees.
+In the following, we would like to plot the evolution of the MCMC chain
+and the trees that were sampled. First, we convert the data from the serialized
+format to a _pureMCMCdata_ format. This is a simple data class that 
+contains the trees and the log probabilities of the trees.
 
 ```{python}
 #| warning: false
@@ -163,6 +163,6 @@ Is it perhaps the true tree?
 yg.compare_trees(last_tree, true_tree)
 ```
 
-Not note that the last tree does not need to be a good tree. SCITE is just likely to spend more iterations exploring more likely trees. Here the last tree just turns out to be a tree with the highest log-probability.
+Now note that the last tree does not need to be a good tree. SCITE is just likely to spend more iterations exploring more likely trees. Here, the last tree just turns out to be a tree with the highest log-probability.
 
- To acutally retrive a mutation tree from the posterior one would have to make a point estimate Maximum A Posteriori (MAP) tree, i.e. sampled the most times.  See SCITE paper for details.
+To acutally retrive a mutation tree from the posterior one would have to make a point estimate Maximum A Posteriori (MAP) tree, i.e. sampled the most times.  See SCITE paper for details.