we learn the name of zuck’s billion-dollar target: andrew tulloch, mira’s chief systems architect, former openai, former meta “the subject of a spirited recruitment campaign” he turned down a potential $1.5 billion

The first thing people do when progressing from low income to lower-middle income in most of the country is buy a car. It’s the single biggest improvement in quality of life.

Diffusing Protein Binders to Intrinsically Disordered Proteins 🚀 New paper from David Baker!🚀 1. A groundbreaking study published in Nature demonstrates a novel method for designing protein binders that can target intrinsically disordered proteins (IDPs) and regions (IDRs) with high affinity and specificity. This approach leverages RFdiffusion, a powerful computational tool that generates binders by freely sampling both target and binding protein conformations from the target sequence alone. 2. The research team successfully created binders for a diverse set of IDPs and IDRs, including amylin, C-peptide, VP48, and BRCA1_ARATH, with dissociation constants (Kd) ranging from 3 to 100 nM. These binders were shown to bind their respective targets in cells, highlighting the potential for therapeutic and diagnostic applications. 3. A key innovation is the ability to design binders without pre-specifying the target geometry, allowing the binder to select a specific conformation from a broad ensemble of possible states. This induced-fit mechanism is particularly effective for IDPs, which lack a single well-defined structure. 4. The study also introduced a two-sided partial diffusion approach to optimize binding affinity, resulting in binders with significantly improved metrics compared to traditional one-sided diffusion methods. This technique allows both the target and binder to adapt their conformations during the design process. 5. For shorter IDRs, the researchers incorporated secondary structure specification into the RFdiffusion model, enabling the design of binders that interact with β-strand conformations of the target. This method significantly increased the efficiency of generating high-affinity binders. 6. The designed binders were validated through crystallography, showing close agreement between the computational models and experimental structures. Additionally, the binders demonstrated high specificity for their intended targets in all-by-all binding experiments. 7. In cellular experiments, the binders colocalized with their full-length targets, confirming their ability to engage targets in a cellular context. Notably, the G3BP1 binder disrupted stress granule formation, while the amylin binder inhibited amyloid fibril formation and dissociated existing fibers. 8. The amylin binder was further shown to enhance the sensitivity of mass spectrometry-based amylin detection and enable lysosomal targeting of amylin monomers and fibrils, demonstrating the potential for improving diagnostic techniques and therapeutic interventions. 9. This work represents a significant advancement in the field of protein design, offering a versatile and powerful approach to targeting IDPs and IDRs. The method has broad implications for understanding and modulating protein interactions in various biological contexts. 💻Code: doi.org/10.5281/zenodo.15453… 📜Paper: nature.com/articles/s41586-0… #ProteinDesign #IDPs #Therapeutics #ComputationalBiology #Innovation