Volume 393 | Issue 6807 | 9 July 2026 — Summary of Science publications, prioritizing content relevant to AI and life sciences.

Note: this issue is comparatively thin on AI-specific research — I verified each item below individually against its DOI and issue tag rather than trust aggregated "contents" search snippets, which repeatedly mixed content from the adjacent 6806/6808 issues. I could not confirm the official cover story through reliable sourcing, so I'm not guessing at one.

🧠 TOP LIFE-SCIENCES STORY — Neuronal Hyperactivity Recruits Antibody-Secreting Cells to Prune Adult Synapses

Research Article · Crowley et al. · DOI: 10.1126/science.adv1219 · see Perspective by Miarka & Prinz, pp. 143–144: DOI 10.1126/science.aei8076

Complement-mediated synaptic pruning (via C1q) is well established in brain development and in neurodegenerative disease, but its role in the otherwise healthy adult brain has been unclear. Using in vivo chemogenetics, this study shows that neuronal hyperactivity in the adult hippocampus triggers C1q deposition at synaptic sites alongside recruitment of antibody-secreting plasma cells — directly implicating the adaptive immune system in activity-dependent synaptic remodeling. The Perspective reframes developmental complement pruning as a constitutive "hyperactivity sensor" active throughout adult life, with implications for epilepsy and cognitive decline research.

Science

📊 TOP AI/METHODOLOGY STORY — A Statistical Test for the Benefits of Personalizing Interventions

Research Article · DOI: 10.1126/science.aeb9506

A methods paper addressing a core tension in medicine, education, and AI-driven services: when is it actually worth investing in a personalized approach versus a cheaper, universal one? The authors propose a statistical framework for making that call rather than defaulting to intuition. Directly applicable to precision-medicine trial design and to any AI system deciding how much to tailor its output per user — a useful methodological reference for Bioinfo.AI's coverage of AI-driven personalization in clinical or research tools.

Science

🧲 Orbital Angular Momentum Generation Survives Structural Disorder

In Other Journals · pp. 158–159 · DOI: 10.1126/science.aek3449

Summarizing Kang et al., this piece covers manganese films ranging from amorphous to single-crystalline, all of which generated orbital angular momentum (OAM) — a property of growing interest as a complement to spin/charge in next-generation computing and sensing hardware — with robustness to disorder that's promising for practical device fabrication. Tangential AI-hardware relevance rather than a core AI/life-sciences story.

Science

🦉 How Much Killing Will It Take to Save the Spotted Owl? — News Feature

News · By Warren Cornwall

An in-depth report on the escalating lethal-control campaign against invasive barred owls in Pacific Northwest old-growth forests, undertaken to protect the native northern spotted owl. Conservation policy content rather than AI/life-sciences research, included as the issue's flagship news story (also the subject of the week's podcast, "Owl wars and the immune system's memory").

Science

Bottom line for Bioinfo.AI: If you want a stronger AI+life-sciences anchor for this week's slot, the neighboring issue — 393/6808 (16 July 2026) — contains a much better fit: a Jennifer Doudna-lab paper (Skopintsev et al., pp. 313–318) using AI inverse-folding models (ESM-IF1) combined with evolution-guided constraints to design entirely new-to-nature CRISPR RNA-guided nucleases ("SynTnpBs"), validated with cryo-EM and gene-editing assays across bacterial, plant, and human cells. I can produce a full summary of 6808 if you'd like — it would likely make a stronger blog post than 6807 on its own.