<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/c0cd25175dea4cc199034cb75f3358ba&quot; frameborder=&quot;0&quot; width=&quot;1280&quot; height=&quot;960&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>960</height><width>1280</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>960</thumbnail_height><thumbnail_width>1280</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/c0cd25175dea4cc199034cb75f3358ba-45f902becda01769.gif</thumbnail_url><duration>265.884</duration><title>Gemma 4 hackathon Solution</title><description>This Loom describes the Cell2Center system for converting transcriptional information from CellXGene data into natural in-frame annotations such as cell-type labels. The author explains that a dataset is processed with a visual pipeline that is run beforehand since it takes time. They mention using models or components referred to as WP Express names and Gemma, including “Luna,” and discuss achieving accurate gene structure interpretation. The discussion concludes by noting the gene structure is highly mixed and that the approach can compress hours of work into minutes.</description></oembed>