Red Light Therapy Moves From Wellness Hype Toward Scientific Legitimacy: Red-light therapy, once dismissed as fringe wellness culture, is gaining scientific credibility as researchers uncover evidence that red and near-infrared light can influence cellular function, particularly through mitochondria. Clinical studies and expert reviews now support its use in a few areas, including oral mucositis, certain ulcers, peripheral neuropathy, hair loss, and some retinal conditions, while early work hints at possible benefits for brain disorders, pain, metabolism, and recovery. At the same time, scientists caution that the field is crowded with exaggerated commercial claims, uneven device quality, and unresolved questions about optimal dosing, delivery, and biological mechanisms. The article also raises a broader possibility: modern indoor life may deprive people of wavelengths humans evolved with, making light not just a treatment issue but a public-health one. (Nature)

Malone Quits Kennedy Vaccine Panel Amid Internal Feud: Robert Malone, a vice chair of the CDC’s vaccine advisory panel under Health Secretary Robert F. Kennedy Jr., has resigned after a public dispute with an HHS spokesperson. Malone said the final trigger was spokesperson Andrew Nixon publicly dismissing Malone’s claim that the Advisory Committee on Immunization Practices might be dissolved and rebuilt again. Malone later retracted that assertion, calling it a miscommunication. His departure comes amid broader turmoil surrounding ACIP, which Kennedy repopulated with allies after firing all 17 previous members last June. A federal judge has since temporarily blocked those appointments and paused the vaccine-policy changes the new panel had made, ruling they were likely unlawful. Malone said the resignation followed months of unpaid work, hostile press, internal conflict, leaks, and what he described as sabotage. (Ars Technica)

Personalized mRNA Vaccines Show Durable Immune Memory in Triple-Negative Breast Cancer: A new Nature paper reports encouraging long-range follow-up for individualized neoantigen mRNA vaccines in triple-negative breast cancer, one of the harder solid tumors to keep from returning. Researchers tracked 14 patients treated after surgery and standard therapy, finding that nearly all mounted strong vaccine-induced T-cell responses against multiple tumor neoantigens. Those responses were not fleeting: they remained functional for years, and 11 patients stayed relapse-free for as long as six years after vaccination. The three recurrences were informative rather than simply discouraging, pointing to weak initial immune response or tumor escape mechanisms such as low MHC class I expression. The significance is less that cancer vaccines are suddenly “solved” than that tailored RNA immunization is showing persistence, biological plausibility, and clinically meaningful signal in a difficult disease setting. (Nature)

INSTALL Points to a Gentler Route for Large-Scale Genome Insertion: One of the more intriguing gene-editing developments this week is INSTALL, a strategy highlighted by Nature for inserting large DNA sequences while sidestepping a major bottleneck: innate immune detection of donor DNA. The system combines mostly single-stranded DNA donors, which are less inflammatory, with a short double-stranded region that still gives recombinase enzymes something to grab onto. In human and mouse cells, the approach reportedly enables insertion of larger genetic cargo without the toxic immune responses that can plague more conventional methods. That matters because many therapeutic ambitions in gene editing require adding substantial chunks of DNA, not just correcting single-letter mutations. If INSTALL holds up across more cell types and disease models, it could expand the practical toolbox for gene replacement and synthetic biology, especially where delivery toxicity has limited ambition. (Nature)

FDA and NIH Push Harder Toward Animal-Test Alternatives: A notable U.S. policy-and-platform story this week came from Fierce Biotech: the FDA and NIH are trying to give non-animal methods more regulatory momentum. FDA officials said the agency’s new guidance is meant to widen flexibility for developers using NAMs, or new approach methodologies, in areas where targets and product classes are well characterized, while still stopping short of abolishing animal studies altogether. At the same time, NIH announced $150 million in Complement-ARIE awards to support seven technology development centers, a data hub, and coordination infrastructure. Priority areas include gynecological, cardiac, neurological, and rare diseases. The broader significance is that this is no longer just an ethical conversation about reducing animal use; it is becoming a federal infrastructure build-out around organoids, data platforms, and validation pathways intended to make human-relevant preclinical evidence more usable in real regulatory settings. (Fierce Biotech)

Merck Buys Terns in a $6.7 Billion Oncology Move: Merck’s $6.7 billion agreement to acquire Terns Pharma is one of the week’s clearest biotech business signals: big pharma remains willing to spend aggressively where it sees late-decade oncology risk. The immediate attraction is TERN-701, an experimental chronic myeloid leukemia treatment that showed a 75% major molecular response rate in an early study among previously treated patients. Analysts cited by Reuters suggest the drug could emerge as a challenger or successor to current standards such as Novartis’s Scemblix. For Merck, the timing is strategic. Keytruda generated more than $30 billion in 2025 and accounts for nearly half the company’s revenue, but patent losses loom later in the decade. This deal fits a broader pattern of pipeline bulking before that cliff arrives. The premium was modest, but the message was not: Merck is still shopping for assets that can matter in oncology at scale. (Reuters)

Gilead Expands Beyond Core Franchises With Ouro Deal: Gilead’s move to buy privately held Ouro Medicines for up to $2.18 billion is smaller than Merck’s Terns acquisition but still significant in what it says about where biotech money is going. Reuters’ report was spare on pipeline detail, but the strategic logic is plain: Gilead is still looking beyond its established strongholds as patent pressure and changing revenue mixes force large drugmakers to refresh their portfolios. In the current biotech climate, deals like this matter not only for the assets acquired but for the signal they send about platform areas investors and pharmaceutical buyers still find compelling. When one of the sector’s major players spends more than $2 billion on a private biotech, it suggests the market for differentiated immunology or adjacent therapeutic pipelines remains active even amid tighter capital discipline. It is another reminder that consolidation remains one of biotech’s defining 2026 storylines. (Reuters)

Protein Sequencing Edges Closer to DNA-Scale Readout: A Stanford-led team has described a protein-sequencing approach that tries to piggyback on one of modern biology’s biggest advantages: the speed and cost efficiency of DNA sequencers. As summarized by Phys.org from a March 18 Nature Biotechnology paper, the researchers developed chemistry that effectively “reverse translates” proteins into DNA-readable information by tagging amino acids with DNA barcodes and positional markers. The idea is to let routine DNA-sequencing platforms decode proteins molecule by molecule. That is a provocative step because proteomics has long lagged genomics in throughput and sensitivity, even though proteins carry much of the functional information that DNA alone cannot capture. The team argues its method could potentially reveal orders of magnitude more molecules from a sample than mass spectrometry typically does, including rare proteins. If that scales, it could reshape single-cell biology, diagnostics, and the study of disease states where protein variation is decisive. (Phys.org)

AI-Guided Base Editing Tool Raises the Bar for Precision: Another gene-editing advance worth watching comes from NUS Medicine, where researchers used AlphaFold3-guided design plus a new “Trinity-Screen” selection system to improve the performance of a base-editing enzyme. According to Phys.org’s report, the screen simultaneously filtered for three qualities inside cells: editing efficiency, low DNA breakage, and low toxicity. The best engineered variants showed as much as 11.8-fold greater editing efficiency than the original enzyme, roughly half the unwanted DNA-break rate seen in earlier high-activity mutants, and up to a 31-fold improvement on a composite performance index. The editors were also smaller than many existing tools, which could matter for packaging them into viral delivery systems used in gene therapy. The story is important because it is not just about stronger editing; it is about making precision, safety, and deliverability part of the same design loop rather than treating them as after-the-fact compromises. (Phys.org)

A Genetic Switch in Wild Rice Could Make Perennial Crops More Practical: Rice research usually lands in agriculture pages, but this week’s perennial-rice story is really a biotechnology and sustainability story. Reporting on a new Science paper, Phys.org described how researchers identified a genomic region called Endless Branches and Tillers 1, or EBT1, containing tandem microRNA genes MIR156B and MIR156C, that helps wild rice keep producing new shoots after flowering. When EBT1 was disabled in wild rice, perennial growth was lost; when introduced into cultivated rice, the plant reset its growth cycle instead of shutting down after seed production. The difference appears tied to epigenetic regulation, with cultivated rice turning the genes off as it matures while wild rice can reactivate them. The practical appeal is obvious: perennial rice could cut yearly replanting costs, labor, and land disturbance. The deeper importance is that a developmental habit long treated as a breeding trait now looks much more tractable at the molecular level. (Phys.org)

Cornell’s Grain-of-Salt Brain Implant Pushes Neural Interfaces Smaller: On the technology side, one of the week’s standout stories is Cornell’s microscale optoelectronic tetherless electrode, or MOTE, an implant so small it can sit on a grain of salt while wirelessly reporting brain activity for more than a year in animals. ScienceDaily’s summary of the Nature Electronics study says the device is about 300 microns long and 70 microns wide, uses safe red and infrared laser beams for power and communication, and transmits neural signals via tiny pulses of infrared light. The hardware includes a semiconductor diode, low-noise amplifier, and optical encoder built at microsystem scale. Miniaturization is the headline, but the larger point is what that size reduction unlocks: chronic recording with less tethering, possible MRI compatibility, and the prospect of adapting similar systems for spinal or other bio-integrated sensing applications. Neural interfaces often advance by incremental engineering; this looks like one of the more meaningful step changes. (ScienceDaily)

Oral Insulin Gets a More Credible Path Forward: A long-running diabetes dream got a fresh boost this week with a ScienceDaily report on Kumamoto University’s peptide-based platform for oral insulin delivery. The core problem has always been brutal: digestive enzymes break insulin down, and the intestine is poor at absorbing it into the bloodstream. The Japanese team’s workaround uses a cyclic peptide, DNP, that can traverse the small intestine and carry insulin with it, either by mixing with zinc-stabilized insulin hexamers or by being chemically linked directly to insulin. In diabetic mouse models, both approaches brought blood sugar down effectively, and the platform achieved about 33% to 41% pharmacological bioavailability relative to injected insulin. That is the kind of number that shifts oral insulin from elegant idea toward plausible product architecture. There is still a long road through larger-animal and human studies, but this looks like one of the more technically credible oral-delivery advances in years. (ScienceDaily)

A Tiny Jellyfish Suggests Biology May Tell Time in More Than One Way: For a broader pure-science pick, Quanta highlighted a striking chronobiology result: a newly described hydrozoan jellyfish appears to keep time without the canonical animal circadian-clock genes. The species, found off Japan, seems to track roughly 20-hour cycles and also uses a separate sunrise-linked timer tied to spawning. That combination is important because standard animal clocks are usually built around familiar genes such as CLOCK, BMAL1, and CRY. Hydrozoans seem to have lost those components, yet this jellyfish still behaves as if it possesses a functioning biological clock. The implication is larger than an odd marine anecdote. It suggests researchers may have been too narrow in defining what counts as a circadian mechanism, and that alternative molecular architectures for timekeeping may be hiding across the tree of life. In science terms, it is a reminder that evolutionary solutions can be stranger—and broader—than our model systems imply. (Quanta Magazine)

IMAGE CREDIT: NASA.