<p>** WARNING** This tutorial is under construction, and we do not advise running it at this point in time.</p>
<p>This tutorial covers the basics of using hdWGCNA to perform <strong>isoform</strong> co-expression network analysis using <ahref="https://www.pacb.com/wp-content/uploads/Application-note-MAS-Seq-for-single-cell-isoform-sequencing.pdf"class="external-link">PacBio MAS-Seq</a> data. Briefly, MAS-Seq is a single-cell RNA-seq protocol that uses long-read sequencing to profile full-length RNA molecules, potentially providing a much greater depth of information about individual RNA splicing isoforms compared to traditional short-read scRNA-seq. This tutorial starts from the typical genes by cells counts matrix, as well as an isoforms by cells counts matrix unique to long-read single-cell approaches like MAS-seq, and we cover how to perform clustering analysis in long-read single-cell data using Seurat. Finally, we demonstrate several different ways to use hdWGCNA for isoform co-expression network analysis. Here we use a MAS-Seq dataset of PBMCs from two donors, sequenced using the PacBio Revio platform. Before working through this tutorial, we recommend becoming familar with the basics of hdWGCNA using our single-cell tutorial.</p>
<divclass="section level2">
<h2id="download-the-tutorial-data">Download the tutorial data<aclass="anchor"aria-label="anchor"href="#download-the-tutorial-data"></a>
</h2>
<p>Here we download the genes and isoforms counts matrices from PacBio for the two replicates using wget. These counts matrices were generated using the<br>
PacBio SMRT Link software, more details can be found at the <ahref="https://isoseq.how/umi/"class="external-link">isoseq documentation website</a>.</p>
<p>Here we download the genes and isoforms counts matrices from PacBio for the two replicates using wget. These counts matrices were generated using the PacBio SMRT Link software, more details can be found at the <ahref="https://isoseq.how/umi/"class="external-link">isoseq documentation website</a>.</p>
<p>You can also <ahref="https://drive.google.com/file/d/1pIINwgMlJX-3db4sLApt5Gi7GkWQElGf/view?usp=share_link"class="external-link">download the unprocessed Seurat object containing both replicates</a> instead of running the “Create a multi-assay Seurat object” section.</p>
@@ -175,7 +174,7 @@ PacBio SMRT Link software, more details can be found at the <a href="https://iso
<divclass="section level3">
<h3id="create-a-multi-assay-seurat-object">Create a multi-assay Seurat object<aclass="anchor"aria-label="anchor"href="#create-a-multi-assay-seurat-object"></a>
</h3>
<p>The MAS-Seq dataset includes both gene and isoform level quantifications, and in this section we will load these matrices for both replicates to create a multi-assay Seurat object. The number of isoforms detected differs for each replicate, and here we have chosen to only include isoforms that were detected in both replicates.</p>
<p>The MAS-Seq dataset includes both gene and isoform level quantifications, and in this section we will load these matrices for both replicates to create a multi-assay Seurat object. The number of isoforms detected differs for each replicate, and here we have chosen to only include isoforms that were detected in both replicates. If you don’t want to run this section, you can <ahref="https://drive.google.com/file/d/1pIINwgMlJX-3db4sLApt5Gi7GkWQElGf/view?usp=share_link"class="external-link">download the unprocessed Seurat object here</a>.</p>
<p>This approach gives us 13217 isoforms for network analysis. Now we run the main steps of co-expression network analysis with hdWGCNA similarly to what we did in the previous section.</p>
<h3id="example-3-major-isoforms-in-one-cell-type">Example 3: Major isoforms in one cell type<aclass="anchor"aria-label="anchor"href="#example-3-major-isoforms-in-one-cell-type"></a>
<h3id="example-3-consensus-wgcna-between-the-two-samples">Example 3: consensus wgcna between the two samples<aclass="anchor"aria-label="anchor"href="#example-3-consensus-wgcna-between-the-two-samples"></a>
metacell_location <spanclass="op">=</span><spanclass="st">'variable'</span>, <spanclass="co"># this will use the same Metacell seurat object that we made before</span>
features <spanclass="op">=</span><spanclass="fu"><ahref="https://rdrr.io/r/base/rev.html"class="external-link">rev</a></span><spanclass="op">(</span><spanclass="va">mods</span><spanclass="op">)</span>
<h3id="bonus-consensus-wgcna-between-the-two-samples">Bonus: consensus wgcna between the two samples:<aclass="anchor"aria-label="anchor"href="#bonus-consensus-wgcna-between-the-two-samples"></a>
</h3>
</div>
<divclass="section level2">
<h2id="isoform-plotting-with-ggtranscript-and-swan">Isoform plotting with ggtranscript and Swan<aclass="anchor"aria-label="anchor"href="#isoform-plotting-with-ggtranscript-and-swan"></a>
