forked from alshedivat/al-folio
-
Notifications
You must be signed in to change notification settings - Fork 0
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Deploying to gh-pages from @ b5ca376 🚀
- Loading branch information
Showing
13 changed files
with
72 additions
and
53 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Large diffs are not rendered by default.
Oops, something went wrong.
Large diffs are not rendered by default.
Oops, something went wrong.
Large diffs are not rendered by default.
Oops, something went wrong.
Large diffs are not rendered by default.
Oops, something went wrong.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,32 @@ | ||
| <!DOCTYPE html> <html> <head> <meta http-equiv="Content-Type" content="text/html; charset=UTF-8"> <meta charset="utf-8"> <meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no"> <meta http-equiv="X-UA-Compatible" content="IE=edge"> <title>Review of the paper Prediction of mechanistic subtypes of Parkinson's using patient derived stem cell models | Xuelong An Wang</title> <meta name="author" content="Xuelong An Wang"> <meta name="description" content="comments on a paper that leverages deep learning to classify cells into Parkinson's subtypes"> <meta name="keywords" content="jekyll, jekyll-theme, academic-website, portfolio-website"> <link href="https://cdn.jsdelivr.net/npm/bootstrap@4.6.1/dist/css/bootstrap.min.css" rel="stylesheet" integrity="sha256-DF7Zhf293AJxJNTmh5zhoYYIMs2oXitRfBjY+9L//AY=" crossorigin="anonymous"> <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/mdbootstrap@4.20.0/css/mdb.min.css" integrity="sha256-jpjYvU3G3N6nrrBwXJoVEYI/0zw8htfFnhT9ljN3JJw=" crossorigin="anonymous"> <link defer rel="stylesheet" href="https://unpkg.com/bootstrap-table@1.21.4/dist/bootstrap-table.min.css"> <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/@fortawesome/fontawesome-free@6.4.0/css/all.min.css" integrity="sha256-HtsXJanqjKTc8vVQjO4YMhiqFoXkfBsjBWcX91T1jr8=" crossorigin="anonymous"> <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/academicons@1.9.1/css/academicons.min.css" integrity="sha256-i1+4qU2G2860dGGIOJscdC30s9beBXjFfzjWLjBRsBg=" crossorigin="anonymous"> <link rel="stylesheet" type="text/css" href="https://fonts.googleapis.com/css?family=Roboto:300,400,500,700|Roboto+Slab:100,300,400,500,700|Material+Icons"> <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jwarby/jekyll-pygments-themes@master/github.css" media="" id="highlight_theme_light"> <link rel="shortcut icon" href="/assets/img/brain-fractal-3.jpg"> <link rel="stylesheet" href="/assets/css/main.css?d41d8cd98f00b204e9800998ecf8427e"> <link rel="canonical" href="https://awxlong.github.io/blog/2023/mechanistic-subtypes-parkinson-copy/"> <link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jwarby/jekyll-pygments-themes@master/native.css" media="none" id="highlight_theme_dark"> <script src="/assets/js/theme.js?96d6b3e1c3604aca8b6134c7afdd5db6"></script> <script src="/assets/js/dark_mode.js?9b17307bb950ffa2e34be0227f53558f"></script> <script src="https://cdn.jsdelivr.net/npm/jquery@3.6.0/dist/jquery.min.js" integrity="sha256-/xUj+3OJU5yExlq6GSYGSHk7tPXikynS7ogEvDej/m4=" crossorigin="anonymous"></script> <script type="text/javascript">window.MathJax={tex:{tags:"ams"}};</script> <script defer type="text/javascript" id="MathJax-script" src="https://cdn.jsdelivr.net/npm/mathjax@3.2.0/es5/tex-mml-chtml.js"></script> <script defer src="https://polyfill.io/v3/polyfill.min.js?features=es6"></script> <script src="/assets/js/distillpub/template.v2.js"></script> <script src="/assets/js/distillpub/transforms.v2.js"></script> <script src="/assets/js/distillpub/overrides.js"></script> </head> <body> <d-front-matter> <script async type="text/json">{ | ||
| "title": "Review of the paper Prediction of mechanistic subtypes of Parkinson's using patient derived stem cell models", | ||
| "description": "comments on a paper that leverages deep learning to classify cells into Parkinson's subtypes", | ||
| "published": "November 2, 2023", | ||
| "authors": [ | ||
| { | ||
| "author": "Xuelong An", | ||
| "authorURL": "", | ||
| "affiliations": [ | ||
| { | ||
| "name": "", | ||
| "url": "" | ||
| } | ||
| ] | ||
| } | ||
|
|
||
| ], | ||
| "katex": { | ||
| "delimiters": [ | ||
| { | ||
| "left": "$", | ||
| "right": "$", | ||
| "display": false | ||
| }, | ||
| { | ||
| "left": "$$", | ||
| "right": "$$", | ||
| "display": true | ||
| } | ||
| ] | ||
| } | ||
| }</script> </d-front-matter> <header> <nav id="navbar" class="navbar navbar-light navbar-expand-sm fixed-top"> <div class="container"> <a class="navbar-brand title font-weight-lighter" href="/"><span class="font-weight-bold">Xuelong </span>An Wang</a> <button class="navbar-toggler collapsed ml-auto" type="button" data-toggle="collapse" data-target="#navbarNav" aria-controls="navbarNav" aria-expanded="false" aria-label="Toggle navigation"> <span class="sr-only">Toggle navigation</span> <span class="icon-bar top-bar"></span> <span class="icon-bar middle-bar"></span> <span class="icon-bar bottom-bar"></span> </button> <div class="collapse navbar-collapse text-right" id="navbarNav"> <ul class="navbar-nav ml-auto flex-nowrap"> <li class="nav-item "> <a class="nav-link" href="/">about</a> </li> <li class="nav-item active"> <a class="nav-link" href="/blog/">blog<span class="sr-only">(current)</span></a> </li> <li class="nav-item "> <a class="nav-link" href="/publications/">research</a> </li> <li class="nav-item "> <a class="nav-link" href="/projects/">projects</a> </li> <li class="nav-item "> <a class="nav-link" href="/cv/">cv</a> </li> <li class="toggle-container"> <button id="light-toggle" title="Change theme"> <i class="fas fa-moon"></i> <i class="fas fa-sun"></i> </button> </li> </ul> </div> </div> </nav> <progress id="progress" value="0"> <div class="progress-container"> <span class="progress-bar"></span> </div> </progress> </header> <div class="post distill"> <d-title> <h1>Review of the paper Prediction of mechanistic subtypes of Parkinson's using patient derived stem cell models</h1> <p>comments on a paper that leverages deep learning to classify cells into Parkinson's subtypes</p> </d-title> <d-byline></d-byline> <d-article> <d-contents> <nav class="l-text figcaption"> <h3>Contents</h3> <div><a href="#brief-summary">Brief summary</a></div> <div><a href="#my-comments-and-future-research-directions">My comments and future research directions</a></div> </nav> </d-contents> <h1 id="brief-summary">Brief summary</h1> <p>The paper by <d-cite key="dsa_2023_prediction"></d-cite> leverages deep learning architectures to solve a pentanary classification task given either a cell’s tabular features or images. The five independent classes are one healthy control and four subtypes of Parkinson’s Disease: familial proteinopathy (SNCA), environmental proteinopathy (\(\alpha\)-Syn oligomer), and two subtypes characterized by different mitochondria dysfunction pathways. These pathologies were chemically induced on stem cells. Fifty-six phenotypical features of them were extracted automatically and recorded as tabular data, along with images of the cells extracted via microscopy. Both data modalities were labeled with one of the five classes.</p> <p>The research team trained separately a dense feedforward neural network (DNN) to classify on the tabular data, as well as a convolutional neural network (CNN) to classify on image data. The test classification accuracy achieved by the DNN reached around 83%, while the CNN 95%.</p> <figure> <img src="/assets/img/parkinson.png" alt="Sorry. Image couldn't load." width="100%" height="auto"> <figcaption id="bottleneck">Two separate models are trained on different datasets on the same task of Parkinson subtype classification. Figure extracted from the original research article at https://www.nature.com/articles/s42256-023-00702-9</figcaption> </figure> <h1 id="my-comments-and-future-research-directions">My comments and future research directions</h1> <p>Generally, in the deep learning literature, it is acknowledged that the usage of DNNs comes at the expense of poor explainability. Despite achieving high classification accuracy, these models are black-boxes. Nonetheless, there are ways to identify what are the features that the neural networks pay the most attention when deciding on a classification label, mainly by looking at its last layer’s activation and tracing back to the input space which input feature is associated to it. In CNNs, the technique employed by the research team is called the ShAP (SHapley Additive exPlanations) method.</p> <p>The authors managed to identify in both the DNN and CNN that the mitochondria, lysosome and the interaction of both features mainly contributed to the classification decisions of both models.</p> <p>One future research direction I am interested is exploring whether by integrating both data sources can improve performance and yield explainability, because the original work trains separate models, trained on different datasets.</p> <p>One source of inspiration is from <d-cite key="li_2023_v1t"></d-cite>, where they integrate image data along with a mouse’s behavioral features to predict its neural responses collected from neural recordings. Another source of inspiration is drawn from concept-bottleneck models <d-cite key="koh_2020_concept"></d-cite>. There, a CNN in charge of processing images doesn’t learn to output a classification label, but instead to output features that are relevant to the image. These features, in turn, are annotations of the image stored in tabular:</p> <figure> <img src="/assets/img/bottleneck.png" alt="Sorry. Image couldn't load." width="100%" height="auto"> <figcaption id="bottleneck">A depiction of the pipeline of a concept-bottleneck model. The first half outputs a set of concepts given an image, which can be learnt from intricate annotations, or metadata, of the image. The second half outputs a classification label. Figure extracted from the original paper </figcaption> </figure> <p>Altogether, with regards to the work by <d-cite key="dsa_2023_prediction"></d-cite>, one interesting extension to their CNN is to have it not predict a Parkinson subtype, but rather learn to predict the cell’s physiological features stored as tabular data given image input. Subsequently, use the features to train a multi-class regressor using standard softmax to output a classification label. The prospect is that this hybrid model can leverage the high accuracy prediction of the CNN, whilst being explainable thanks to the logistic regressor.</p> <p>As a further improvement, we can use a <a href="https://arxiv.org/abs/2210.11394" rel="external nofollow noopener" target="_blank">Slot Transformer</a> instead of the CNN with the hope of learning a disentangled representation given the image with its annotations. However, the architecture will be more computationally expensive. A pretrained Slot Transformer that already learnt to disentangle CLEVR-Scenes may be more powerful than training it from scratch.</p> </d-article> <d-appendix> <d-footnote-list></d-footnote-list> <d-citation-list></d-citation-list> </d-appendix> <d-bibliography src="/assets/bibliography/deep-med.bib"></d-bibliography><div id="disqus_thread" style="max-width: 800px; margin: 0 auto;"> <script type="text/javascript">var disqus_shortname="al-folio",disqus_identifier="/blog/2023/mechanistic-subtypes-parkinson-copy",disqus_title="Review of the paper Prediction of mechanistic subtypes of Parkinson's using patient derived stem cell models";!function(){var e=document.createElement("script");e.type="text/javascript",e.async=!0,e.src="//"+disqus_shortname+".disqus.com/embed.js",(document.getElementsByTagName("head")[0]||document.getElementsByTagName("body")[0]).appendChild(e)}();</script> <noscript>Please enable JavaScript to view the <a href="http://disqus.com/?ref_noscript" rel="external nofollow noopener" target="_blank">comments powered by Disqus.</a> </noscript> </div> </div> <footer class="fixed-bottom"> <div class="container mt-0"> © Copyright 2023 Xuelong An Wang. Powered by <a href="https://jekyllrb.com/" target="_blank" rel="external nofollow noopener">Jekyll</a> with <a href="https://github.com/alshedivat/al-folio" rel="external nofollow noopener" target="_blank">al-folio</a> theme. Hosted by <a href="https://pages.github.com/" target="_blank" rel="external nofollow noopener">GitHub Pages</a>. Photos from <a href="https://unsplash.com" target="_blank" rel="external nofollow noopener">Unsplash</a>. </div> </footer> <script src="https://cdn.jsdelivr.net/npm/bootstrap@4.6.1/dist/js/bootstrap.bundle.min.js" integrity="sha256-fgLAgv7fyCGopR/gBNq2iW3ZKIdqIcyshnUULC4vex8=" crossorigin="anonymous"></script> <script src="https://cdn.jsdelivr.net/npm/mdbootstrap@4.20.0/js/mdb.min.js" integrity="sha256-NdbiivsvWt7VYCt6hYNT3h/th9vSTL4EDWeGs5SN3DA=" crossorigin="anonymous"></script> <script type="text/javascript">function progressBarSetup(){"max"in document.createElement("progress")?(initializeProgressElement(),$(document).on("scroll",function(){progressBar.attr({value:getCurrentScrollPosition()})}),$(window).on("resize",initializeProgressElement)):(resizeProgressBar(),$(document).on("scroll",resizeProgressBar),$(window).on("resize",resizeProgressBar))}function getCurrentScrollPosition(){return $(window).scrollTop()}function initializeProgressElement(){let e=$("#navbar").outerHeight(!0);$("body").css({"padding-top":e}),$("progress-container").css({"padding-top":e}),progressBar.css({top:e}),progressBar.attr({max:getDistanceToScroll(),value:getCurrentScrollPosition()})}function getDistanceToScroll(){return $(document).height()-$(window).height()}function resizeProgressBar(){progressBar.css({width:getWidthPercentage()+"%"})}function getWidthPercentage(){return getCurrentScrollPosition()/getDistanceToScroll()*100}const progressBar=$("#progress");window.onload=function(){setTimeout(progressBarSetup,50)};</script> </body> </html> |
Oops, something went wrong.