Skip to content

Commit

Permalink
Updated README.md
Browse files Browse the repository at this point in the history
  • Loading branch information
@EdRey05
EdRey05 authored Jan 15, 2024
1 parent 55e6892 commit eecb091
Showing 1 changed file with 85 additions and 52 deletions.
137 changes: 85 additions & 52 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6,16 +6,15 @@
<div class="left-column-35">
<p><strong> Summary </strong></p>
<p class="justify-text"> During my graduate studies, I helped two bachelor's students working in my research group to do a small
project analyzing survival data from a breast cancer clinical trial. The aim was to evaluate a possible association between
low <strong>gene expression</strong> of two genes at once with <strong>increased patient survival</strong>. <br><br>
project analyzing survival data from a breast cancer clinical trial. The aim was to assess any <strong>potential correlation</strong>
between <strong>higher patient survival</strong> and <strong>low expression of two genes simultaneously</strong>. <br><br>
For this purpose, we retrieved data publicly available in <strong><a href="https://www.cbioportal.org/">cBioPortal</a></strong>,
and generated <strong>Kaplan-Meier survival curves</strong>. We focused on the <strong>METABRIC</strong> dataset containing
information for over <strong>2,500 patients</strong> and I generated tools in <strong>Jupyter/Colab notebooks</strong> and a
<strong>Streamlit app (see GIF -->)</strong>) to automate the creation of these plots with <strong>Python</strong>. <br><br>
At the end of the project, we were able to identify <strong><10 gene pairs</strong> showing the behavior of interest. That
<strong>Streamlit app (see GIF --></strong>) to automate the creation of these plots with <strong>Python</strong>. <br><br>
At the end of the project, we were able to identify <strong>less than 10 gene pairs</strong> showing the behavior of interest. That
information was used in combination with other data from different techniques (in-silico and wet-lab) to prioritize further
studies evaluating the effect of inhibtion of those genes in cancer cell models.
</p>
studies evaluating the effect of inhibtion of those genes in cancer cell models. </p>
</div>
<div class="right-column-65">
<video width="100%" height="auto" autoplay loop muted><source src="Images_GIFs_Videos/Preview_003.mp4" type="video/mp4"></video>
Expand All @@ -30,16 +29,17 @@
</strong>). </li>
<li class="justify-text">Each plot required to divide the dataset into <strong>4 groups</strong> to generate <strong>4 survival
curves</strong> (expression: <strong>low-low, low-high, high-low, high-high</strong>). </li>
<li class="justify-text">The clinical data (<strong>survival times and status</strong>) and the RNA Seq expression data were in
different datasets with different structure, so pre-processing to both was required before we could map
the patient IDs. </li>
<li class="justify-text">The clinical data (<strong>survival times and status</strong>) and the <strong>RNA Seq expression</strong>
data were in different datasets that have different structure, so pre-processing to both of them was required
before we could map the patient IDs. </li>
<li class="justify-text">We needed to screen all the plots generated but keep only the ones where the <strong>low-low curve</strong>
was higher than the others, and retrieve relevant data such as <strong>CIs and time to 50% survival</strong>
to complement our analysis. </li>
<li class="justify-text">Since each clinical trial reports the data in a different way and not all have RNA Seq data, we chose the
best possible option for breast cancer (<strong><a href="https://www.cbioportal.org/study/summary?id=brca_metabric">METABRIC</a></strong>). </li>
<li class="justify-text">In order to reuse our code for other breast cancer datasets or even different cancer types, we needed to
generalize the workflow as much as possible and make tools for reproducibility and automation. </li>
<li class="justify-text">In order to reuse our code for other breast cancer datasets or even different cancer types, we <strong>needed
to generalize the workflow</strong> as much as possible and <strong>make tools for reproducibility and
automation</strong>. </li>
</ul>
</div>
<div class="right-column-50">
Expand All @@ -48,15 +48,17 @@
<li class="justify-text">I learned how to use the <strong>KaplanMeierFitter</strong> module from the <strong>lifelines</strong> python
library to generate KM plots. </li>
<li class="justify-text">I first generated a <strong>Google Colab notebook</strong> that was dataset-specific to produce batches of
<strong>40-50</strong> plots. This exclusively makes 4 groups from the original dataset based on RET and one
other gene, which required manual editing of the code to write down all 40-50 names of genes (<a href="https://user-images.githubusercontent.com/62916582/204424020-bae3613c-bf10-4a3b-9d50-beaf50ca8eee.gif" target="_blank">View tool</a>). </li>
<strong>40-50</strong> plots. This <strong>exclusively makes 4 groups</strong> from the original dataset based
on the expression of RET and one other gene, which required to manually write in the code all 40-50 names of the
other gene (<a href="https://user-images.githubusercontent.com/62916582/204424020-bae3613c-bf10-4a3b-9d50-beaf50ca8eee.gif" target="_blank">View tool</a>). </li>
<li class="justify-text">Then, I found a way to generalize some steps and created a <strong>Jupyter notebook</strong> that used
<strong>ipywidgets</strong> to interactively get user inputs, allowing dynamic selection of any columns
to divide the dataset into 2 or more groups and re-plotting curves (<a href="https://github.com/EdRey05/Resources_for_Mulligan_Lab/blob/de82796fe821b96c18ab0709018c02c3b02aba92/Tutorials/Preview_Interactive_KM.gif" target="_blank">View tool</a>). </li>
<li class="justify-text">Finally, I discovered <strong>Streamlit</strong> and adapted my interactive notebook to a data app (GIF above)
that used a similar code approach but with more interactivy, improved outputs and better user experience. </li>
<li class="justify-text">Although the app works for several different datasets, I noticed high variability in the formatting of clinical
trial data, and constantly try to improve the app to generalize it more.
<strong>ipywidgets</strong> to <strong>interactively get user inputs</strong>, allowing dynamic selection of
<strong>any measured variable</strong> to divide the dataset into <strong>2 or more groups</strong> and
re-plotting curves easily (<a href="https://github.com/EdRey05/Resources_for_Mulligan_Lab/blob/de82796fe821b96c18ab0709018c02c3b02aba92/Tutorials/Preview_Interactive_KM.gif" target="_blank">View tool</a>). </li>
<li class="justify-text">Finally, I discovered <strong>Streamlit</strong> and adapted my interactive notebook to a <strong>data app</strong>
(GIF above) that used a similar approach but has <strong>more interactivy, improved outputs and better user experience</strong>. </li>
<li class="justify-text">Although the app works well for several datasets, I noticed <strong>high variability in the formatting of clinical
trial data</strong>, and try to improve my app to generalize it more!. </li>
</ul>
</div>
</div>
Expand All @@ -71,35 +73,51 @@
<div class="two-columns">
<div class="left-column-35">
<p><strong> Summary </strong></p>
<p class="justify-text"> During my graduate studies, I performed numerous microscopy experiments. These experiments required to acquire sufficient
images (of cancer cells), process them and analyze them. When I was setting up a new technique in our research group
called <strong>Proximity Ligation Assay - PLA</strong> to evaluate protein-protein interactions, I was able to use a
cell imager (EVOS M7000 - Thermo Fisher) that has automation capabilities. I automated the acquisition of hundreds of
fields of view (big images with many cells), and wrote scripts in <code>Jython</code> (Python wrapper for Java) to
automate the image pre-processing and analysis in <strong>ImageJ/Fiji</strong>. The final output of my series of scripts
was a csv file with the quantification result for each image of individual cells (big images where cropped into many
smaller). Also, to validate the quantification results I designed a tool to merge all outputs into a summary Power Point
presentation. I was able to automate the creation of slides, define the layout and the items to insert into each slide
by using the <code>python-pptx</code> library. I created a tool in Jupyter/Colab notebook version and then a Streamilt
app that does the exact same thing but provides a better user interface and additional information.
<p class="justify-text"> During my graduate studies, I performed numerous <strong>fluorescence microscopy experiments</strong>. These
experiments required to acquire sufficient images of individual cancer cells, process them and analyze them. Typical analysis of my
experiments would involve <strong>co-localization</strong> between signals on different channels, or <strong>object/particle detection
and counting</strong>. <br><br>
I carried out <strong>Proximity Ligation Assay (PLA)</strong> experiments to evaluate <strong>protein-protein interactions</strong>
using a cell imager (<strong>EVOS M7000 - Thermo Fisher</strong>) that has automated imaging features. I automated the acquisition of
hundreds of images, and wrote scripts in <strong>Jython</strong> (Python wrapper for Java) to automate the image pre-processing and
analysis in <strong>ImageJ/Fiji</strong>. The final output of my series of scripts was a <strong>csv file with the quantification
result for each individual cell</strong> and the <strong>cropped images with their object mask image</strong> (shows colored blobs
which are particles both detected and counted). <br><br>
To <strong>validate the quantification results</strong> before proceeding to statistical analysis, I designed a tool to <strong>consolidate
all the outputs</strong> into a summary Power Point presentation. I automated the creation of slides with a <strong>customized layout
</strong> and insertion of fluorescence + object mask <strong>image pairs with their names and quantification results</strong> using
the <strong>python-pptx</strong> library. <br><br>
I created a tool first in the form of a <strong>Google Colab notebook</strong> and then as a <strong>Streamilt data app</strong> which
helped me analyze outputs for <strong>almost 10,000 images</strong>. With this tool, I was able to easily <strong>compare two quantification
methods, test different pre-processing and object detection parameters, and fully optimize the whole workflow for each experiment</strong>.
</p>
</div>
<div class="right-column-65">
<img src="Images_GIFs_Videos/Preview_002.gif" alt="Streamlit Projects 002 GIF" />
</div>
</div>
<p class="justify-text"><strong> Problem </strong></p>
<p class="justify-text"> Some description here </p>
<p class="justify-text"><strong> Solution </strong></p>
<p class="justify-text"> Some description here </p>
<div class="two-columns">
<div class="left-column-50">
<p><strong> Sample cell images and figure </strong></p>
<img src="Images_GIFs_Videos/Preview_002_figure1.jpg" alt="002_Sample_data_figure" />
<p class="justify-text"><strong> Problem </strong></p>
<ul>
<li class="justify-text"> A. </li>
<li class="justify-text"> B. </li>
<li class="justify-text"> C. </li>
<li class="justify-text"> D. </li>
<li class="justify-text"> E. </li>
<li class="justify-text"> F. </li>
</ul>
</div>
<div class="right-column-50">
<p><strong> First tool </strong></p>
<img src="https://user-images.githubusercontent.com/62916582/204415085-cc39bb7c-904e-487c-a16d-0d894c1e3249.gif" alt="002_first_pre_tool" />
<p class="justify-text"><strong> Solution </strong></p>
<ul>
<li class="justify-text"> A. </li>
<li class="justify-text"> B. </li>
<li class="justify-text"> C. </li>
<li class="justify-text"> D. </li>
<li class="justify-text"> E. </li>
<li class="justify-text"> F. </li>
</ul>
</div>
</div>
<p><strong> Read the instructions and watch another demo of the Streamlit app here: <a href="https://github.com/EdRey05/Streamlit_projects/tree/main/002_Automated_PPTX_PLA">Demo_PPTX_PLA</a></strong></p>
Expand All @@ -120,18 +138,28 @@
<img src="Images_GIFs_Videos/Preview_001.gif" alt="Streamlit Projects 001 GIF" />
</div>
</div>
<p class="justify-text"><strong> Problem </strong></p>
<p class="justify-text"> Some description here </p>
<p class="justify-text"><strong> Solution </strong></p>
<p class="justify-text"> Some description here </p>
<div class="two-columns">
<div class="left-column-50">
<p><strong> DepMap website showing the constant updates to the datasets </strong></p>
<img src="Images_GIFs_Videos/Preview_001_DepMap_website.jpg" alt="001_DepMap_website" />
<p class="justify-text"><strong> Problem </strong></p>
<ul>
<li class="justify-text"> A. </li>
<li class="justify-text"> B. </li>
<li class="justify-text"> C. </li>
<li class="justify-text"> D. </li>
<li class="justify-text"> E. </li>
<li class="justify-text"> F. </li>
</ul>
</div>
<div class="right-column-50">
<p><strong> First tool (used CCLE data from CBioPortals = DepMap 19Q1) </strong></p>
<img src="https://user-images.githubusercontent.com/62916582/204422004-47fe5726-d92d-4193-bc6a-ea30b3a93cc1.gif" alt="001_first_pre_tool" />
<p class="justify-text"><strong> Solution </strong></p>
<ul>
<li class="justify-text"> A. </li>
<li class="justify-text"> B. </li>
<li class="justify-text"> C. </li>
<li class="justify-text"> D. </li>
<li class="justify-text"> E. </li>
<li class="justify-text"> F. </li>
</ul>
</div>
</div>
<p><strong> Read the instructions and watch another demo of the Streamlit app here: <a href="https://github.com/EdRey05/Streamlit_projects/tree/main/001_RNA_expression_DepMap">Demo_RNA_DepMap</a></strong></p>
Expand All @@ -144,14 +172,19 @@
<details><summary markdown="span"> Expand this to read more...</summary>
<div class="two-columns">
<div class="left-column-35">
<p class="justify-text"> If you have a Github account you can create a Codespace with all the requirements to
run my apps. You only have to log into you account, click on the following button, create your Codespace (we
all have 60h of free usage per month!), and follow the instructions. </p>
<a href="https://codespaces.new/EdRey05/Streamlit_projects?quickstart=1" target="_blank">
<img src="https://github.com/codespaces/badge.svg" alt="Open in GitHub Codespaces">
</a>
<br>
<p class="justify-text"> If you have a Github account, you can create a <strong>Github Codespace</strong> with all the requirements to
run my apps. You only have to log into you account, click on the button below, create your Codespace (<strong>we all have 60h of
free usage per month!</strong>), and follow the instructions in this video-->.<br>
*Note that due to size limits, I did everything quickly but added notes so pause, read and see where I clicked! <br><br></p>
<div class="center-text">
<a href="https://codespaces.new/EdRey05/Streamlit_projects?quickstart=1" target="_blank">
<img src="https://github.com/codespaces/badge.svg" alt="Open in GitHub Codespaces">
</a>
</div>
</div>
<div class="right-column-65">
<br>
<video width="100%" height="auto" controls><source src="Images_GIFs_Videos/Demo_Codespaces.mp4" type="video/mp4"></video>
</div>
</div>
Expand Down

0 comments on commit eecb091

Please sign in to comment.