<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
        <title>Columbia NYC: Judging Session</title>
        <link>https://video.igem.org/videos/watch/433091c4-8c66-4d3b-8012-47118a6a4a40</link>
        <description>Stroke remains a leading cause of disability and death worldwide, with therapeutic efficacy declining rapidly post-onset. Although circulating biomarkers offer early diagnosis, the design of high-affinity biosensing binders remains constrained by the opacity of generative protein models. These models operate as black boxes, which obscures the features that drive high affinity and specificity. SENTINEL addresses this bottleneck by introducing a mechanistically-interpretable framework for de novo protein binder engineering. By integrating Sparse Autoencoders with ESM-based protein models, we extract human-interpretable features associated with molecular recognition to steer towards high-affinity binding. We benchmark SENTINEL against leading design architectures and, as proof-of-concept, validate candidate binder pairs for clinically relevant stroke biomarkers (UCH-L1, VILIP-1, and B-FABP) through split reporter survival assays. Through coupling resolved AI design principles and synthetic biology, SENTINEL aims to establish mechanistic interpretability as a universal accelerant for rational protein binder design—adaptable to any disease where early detection saves lives.</description>
        <lastBuildDate>Sat, 21 Nov 2026 11:35:01 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>PeerTube - https://video.igem.org</generator>
        <image>
            <title>Columbia NYC: Judging Session</title>
            <url>https://video.igem.org/client/assets/images/icons/icon-1500x1500.png</url>
            <link>https://video.igem.org/videos/watch/433091c4-8c66-4d3b-8012-47118a6a4a40</link>
        </image>
        <copyright>All rights reserved, unless otherwise specified in the terms specified at https://video.igem.org/about and potential licenses granted by each content's rightholder.</copyright>
        <atom:link href="https://video.igem.org/feeds/video-comments.xml?videoId=433091c4-8c66-4d3b-8012-47118a6a4a40" rel="self" type="application/rss+xml"/>
    </channel>
</rss>