Built site for gh-pages

.nojekyll

@@ -1 +1 @@
-4e7bb5c2
+52c18861

01-processing.html

@@ -373,7 +373,7 @@ <h1 class="title">
 <span><span class="kw">if</span> <span class="op">(</span><span class="va">use_mc</span><span class="op">)</span> <span class="op">{</span></span>
 <span>  <span class="va">nCores</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/Extremes.html">max</a></span><span class="op">(</span><span class="fu">parallel</span><span class="fu">::</span><span class="fu"><a href="https://rdrr.io/r/parallel/detectCores.html">detectCores</a></span><span class="op">(</span><span class="op">)</span><span class="op">/</span><span class="fl">2</span>, <span class="fl">1</span><span class="op">)</span></span>
 <span><span class="op">}</span> <span class="kw">else</span> <span class="op">{</span></span>
-<span>  <span class="va">nCores</span> <span class="op">&lt;-</span> <span class="fl">2</span></span>
+<span>  <span class="va">nCores</span> <span class="op">&lt;-</span> <span class="fl">1</span></span>
 <span><span class="op">}</span></span>
 <span><span class="va">BPPARAM</span> <span class="op">&lt;-</span> <span class="fu">simpleSeg</span><span class="fu">:::</span><span class="fu"><a href="https://sydneybiox.github.io/simpleSeg/reference/generateBPParam.html">generateBPParam</a></span><span class="op">(</span><span class="va">nCores</span><span class="op">)</span></span>
 <span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/theme_get.html">theme_set</a></span><span class="op">(</span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/ggtheme.html">theme_classic</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span></span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
@@ -404,7 +404,7 @@ <h1 class="title">
 <span></span>
 <span><span class="co"># assign metadata columns</span></span>
 <span><span class="fu">mcols</span><span class="op">(</span><span class="va">images</span><span class="op">)</span> <span class="op">&lt;-</span> <span class="fu">S4Vectors</span><span class="fu">::</span><span class="fu"><a href="https://rdrr.io/pkg/S4Vectors/man/DataFrame-class.html">DataFrame</a></span><span class="op">(</span>imageID <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/names.html">names</a></span><span class="op">(</span><span class="va">images</span><span class="op">)</span><span class="op">)</span></span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
-<p>Time for this code chunk to run: 78.31 seconds</p>
+<p>Time for this code chunk to run with 40 cores: 73.99 seconds</p>
 </div>
 <p>When reading the image channels directly from the names of the TIFF images, they will often need to be cleaned for ease of downstream processing. The channel names can be accessed from the <code>CytoImageList</code> object using the <code>channelNames</code> function.</p>
 <div class="cell">
@@ -459,7 +459,7 @@ <h1 class="title">
 <span>                   tissue <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"panCK"</span>, <span class="st">"CD45"</span>, <span class="st">"HH3"</span><span class="op">)</span>,</span>
 <span>                   cores <span class="op">=</span> <span class="va">nCores</span></span>
 <span>                   <span class="op">)</span></span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
-<p>Time for this code chunk to run: 47.92 seconds</p>
+<p>Time for this code chunk to run with 40 cores: 44.22 seconds</p>
 </div>
 <section id="visualise-separation" class="level3" data-number="4.2.1"><h3 data-number="4.2.1" class="anchored" data-anchor-id="visualise-separation">
 <span class="header-section-number">4.2.1</span> Visualise separation</h3>
@@ -505,7 +505,7 @@ <h1 class="title">
 <span>           display <span class="op">=</span> <span class="st">"single"</span>,</span>
 <span>           colour <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/list.html">list</a></span><span class="op">(</span>HH3 <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"black"</span>,<span class="st">"blue"</span><span class="op">)</span>,</span>
 <span>                         CD31 <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"black"</span>, <span class="st">"red"</span><span class="op">)</span>,</span>
-<span>                         FX111A <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"black"</span>, <span class="st">"green"</span><span class="op">)</span> <span class="op">)</span>,</span>
+<span>                         FX111A <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="st">"black"</span>, <span class="st">"green"</span><span class="op">)</span><span class="op">)</span>,</span>
 <span>           legend <span class="op">=</span> <span class="cn">NULL</span>,</span>
 <span>           bcg <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/list.html">list</a></span><span class="op">(</span></span>
 <span>             HH3 <span class="op">=</span> <span class="fu"><a href="https://rdrr.io/r/base/c.html">c</a></span><span class="op">(</span><span class="fl">1</span>, <span class="fl">1</span>, <span class="fl">2</span><span class="op">)</span>,</span>
@@ -519,14 +519,24 @@ <h1 class="title">
 </div>
 </div>
 </div>
-<div class="question">
+<div class="callout callout-style-default callout-tip callout-titled" title="What to look for and change to obtain an ideal segmentation">
+<div class="callout-header d-flex align-content-center">
+<div class="callout-icon-container">
+<i class="callout-icon"></i>
+</div>
+<div class="callout-title-container flex-fill">
+What to look for and change to obtain an ideal segmentation
+</div>
+</div>
+<div class="callout-body-container callout-body">
 <p><strong>Critical thinking</strong></p>
 <ol type="1">
-<li>Is there any information we’re not capturing with this segmentation?</li>
-<li>What parameters might you change to improve the segmentation?</li>
-<li>What are some intrinsic limitations of the simpleSeg method?</li>
+<li>Does the segmentation capture the full nucleus? If not, perhaps you need to try a different transform to improve the thresholding of the nuclei marker. You could also try using <code>pca = TRUE</code> which will borrow information across the markers to help find the nuclei.</li>
+<li>How much of the cell body is the segmentation missing? Try increasing the dilation around the nucleus by setting <code>discSize = 7</code>.</li>
+<li>Are the segmentations capturing neighbouring cells? Try decreasing the dilation to limit lateral spillover of marker signal by setting <code>discSize = 2</code>.</li>
 </ol>
 </div>
+</div>
 <p>Here, we can see that our segmentation mask has done a good job of capturing the CD31 signal, but perhaps not such a good job of capturing the FXIIIA signal, which often lies outside of our dilated nuclear mask. This suggests that we might need to increase the <code>discSize</code> or other parameters of <code>simpleSeg</code>.</p>
 <p>In particular, the <code>cellBody</code> and <code>watershed</code> parameters can strongly influence the way cells are segmented using <code>simpleSeg</code>. We have provided further details on how the user may specify cell body identification and watershedding in the tables below.</p>
 <p>As <code>simpleSeg</code> is a nuclei-based dilation method, it suffers from tissues where cells might be multi-nucleated, or where cells have non-circular or elliptical morphologies. For tissues where you might expect these cells, it may be preferable to choose a different segmentation method.</p>
@@ -627,7 +637,7 @@ <h1 class="title">
 spatialCoords names(2) : x y
 imgData names(1): sample_id</code></pre>
 </div>
-<p>Time for this code chunk to run: 80.61 seconds</p>
+<p>Time for this code chunk to run with 40 cores: 73.95 seconds</p>
 </div>
 <p>So far, we have obtained our raw TIFF images, performed cell segmentation to isolate individual cells, and then stored our data as a <code>SpatialExperiment</code> object. We can now move on to quality control, data transformation, and normalisation to address batch effects.</p>
 </section><section id="sessioninfo" class="level2" data-number="4.4"><h2 data-number="4.4" class="anchored" data-anchor-id="sessioninfo">
@@ -689,7 +699,7 @@ <h1 class="title">
  [52] tidyselect_1.2.1        yaml_2.3.10             abind_1.4-8            
  [55] viridis_0.6.5           codetools_0.2-20        curl_6.0.1             
  [58] lattice_0.22-6          tibble_3.2.1            KEGGREST_1.46.0        
- [61] shiny_1.9.1             withr_3.0.2             evaluate_1.0.1         
+ [61] shiny_1.10.0            withr_3.0.2             evaluate_1.0.1         
  [64] polyclip_1.10-7         Biostrings_2.74.0       filelock_1.0.3         
  [67] BiocManager_1.30.25     pillar_1.9.0            generics_0.1.3         
  [70] sp_2.1-4                RCurl_1.98-1.16         BiocVersion_3.20.0     

02-quality_control.html

@@ -364,7 +364,7 @@ <h1 class="title">
 <span><span class="kw">if</span> <span class="op">(</span><span class="va">use_mc</span><span class="op">)</span> <span class="op">{</span></span>
 <span>  <span class="va">nCores</span> <span class="op">&lt;-</span> <span class="fu"><a href="https://rdrr.io/r/base/Extremes.html">max</a></span><span class="op">(</span><span class="fu">parallel</span><span class="fu">::</span><span class="fu"><a href="https://rdrr.io/r/parallel/detectCores.html">detectCores</a></span><span class="op">(</span><span class="op">)</span><span class="op">/</span><span class="fl">2</span>, <span class="fl">1</span><span class="op">)</span></span>
 <span><span class="op">}</span> <span class="kw">else</span> <span class="op">{</span></span>
-<span>  <span class="va">nCores</span> <span class="op">&lt;-</span> <span class="fl">2</span></span>
+<span>  <span class="va">nCores</span> <span class="op">&lt;-</span> <span class="fl">1</span></span>
 <span><span class="op">}</span></span>
 <span><span class="va">BPPARAM</span> <span class="op">&lt;-</span> <span class="fu">simpleSeg</span><span class="fu">:::</span><span class="fu"><a href="https://sydneybiox.github.io/simpleSeg/reference/generateBPParam.html">generateBPParam</a></span><span class="op">(</span><span class="va">nCores</span><span class="op">)</span></span>
 <span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/theme_get.html">theme_set</a></span><span class="op">(</span><span class="fu"><a href="https://ggplot2.tidyverse.org/reference/ggtheme.html">theme_classic</a></span><span class="op">(</span><span class="op">)</span><span class="op">)</span></span></code><button title="Copy to Clipboard" class="code-copy-button"><i class="bi"></i></button></pre></div>
@@ -399,13 +399,23 @@ <h1 class="title">
 </div>
 </div>
 </div>
-<div class="question">
+<div class="callout callout-style-default callout-note callout-titled">
+<div class="callout-header d-flex align-content-center">
+<div class="callout-icon-container">
+<i class="callout-icon"></i>
+</div>
+<div class="callout-title-container flex-fill">
+Note
+</div>
+</div>
+<div class="callout-body-container callout-body">
 <p><strong>What we’re looking for</strong></p>
 <ol type="1">
-<li>Can we clearly identify the peaks which constitute CD3+ and CD3- cells?</li>
-<li>Do they appear to be largely consistent across images?</li>
+<li>Do the CD3+ and CD3- peaks clearly separate out in the density plot? To ensure that downstream clustering goes smoothly, we want our cell type specific markers to show 2 distinct peaks representing our CD3+ and CD3- cells. If these</li>
+<li>Are our CD3+ and CD3- peaks consistent across our images? We want to make sure that our density plots for CD3 are largely the same across images so that a CD3+ cell in 1 image is equivalent to a CD3+ cell in another image.</li>
 </ol>
 </div>
+</div>
 <p>Here, we can see that the intensities are very clearly skewed, and it is difficult to distinguish a CD3- cell from a CD3+ cell. Further, we can clearly see some image-level batch effect, where across images, the intensity peaks differ drastically.</p>
 <p>Another method of visualising batch effect is using a dimensionality reduction technique and visualising how the images separate out on a 2D plot. If no batch effect is expected, we should see the images largely overlap with each other.</p>
 <div class="cell">
@@ -514,13 +524,23 @@ <h1 class="title">
 </div>
 </div>
 </div>
-<div class="question">
-<p><strong>What we’re looking for</strong></p>
+<div class="callout callout-style-default callout-note callout-titled">
+<div class="callout-header d-flex align-content-center">
+<div class="callout-icon-container">
+<i class="callout-icon"></i>
+</div>
+<div class="callout-title-container flex-fill">
+Note
+</div>
+</div>
+<div class="callout-body-container callout-body">
+<p><strong>Questions revisited</strong></p>
 <ol type="1">
-<li>Can we see that distinct cell types cluster together?</li>
-<li>Can we see the images don’t cluster out separately?</li>
+<li>Do the CD3+ and CD3- peaks clearly separate out in the density plot? If not, we can try optimizing the transform if the distribution looks heavily skewed.</li>
+<li>Are our CD3+ and CD3- peaks consistent across our images? We can try to be more stringent in our normalization, such as by removing the 1st PC (<code>method = c(..., "PC1")</code>) or scaling the values for all images between 0 and 1 (<code>method = c(..., "minMax")</code>).</li>
 </ol>
 </div>
+</div>
 <p>Here, we can see that the normalised data appears more bimodal, and we can clearly observe a CD3+ peak at 5.00, and a CD3- peak at around 3.00. Image-level batch effects also appear to have been mitigated.</p>
 <p>We can also visualise the effect of normalisation on the UMAP, which shows that the images now overlap with each other to a much greater extent.</p>
 <div class="cell">
@@ -603,7 +623,7 @@ <h1 class="title">
  [25] plotly_4.10.4           mime_0.12               lifecycle_1.0.4        
  [28] pkgconfig_2.0.3         rsvd_1.0.5              Matrix_1.7-1           
  [31] R6_2.5.1                fastmap_1.2.0           GenomeInfoDbData_1.2.13
- [34] shiny_1.9.1             digest_0.6.37           colorspace_2.1-1       
+ [34] shiny_1.10.0            digest_0.6.37           colorspace_2.1-1       
  [37] AnnotationDbi_1.68.0    irlba_2.3.5.1           RSQLite_2.3.9          
  [40] beachmat_2.22.0         labeling_0.4.3          filelock_1.0.3         
  [43] fansi_1.0.6             nnls_1.6                httr_1.4.7