Archive
1250 articles, sorted by publication time.
After scaling back research and development, LifeMine has secured funding from Gates, Bezos, RA Capital, and others, and is preparing to use AI to revisit the chemical potential of 100,000 fungal strains. Its first drug has entered a trial in healthy participants, but a long road of validation remains before it can be used in transplantation.
The research team tracked more than 87,000 lung T cells with clues to pathogen recognition and found that some tissue-resident memory clones can persist for months or even years; the results also underscore that blood tests alone may miss the key immunity established by vaccines in the respiratory tract.
SÉMIL interprets the tumor’s invasive front using routine pathology slides and textual descriptions and has been validated across three independent patient cohorts. It could strengthen postoperative risk stratification, but prospective clinical testing is still needed before it can influence chemotherapy decisions.
The research team combined electrocardiographic, electromyographic, and skin-surface mechanical changes into a composite waveform, aiming to reduce wearable-device circuitry and energy consumption. Blood pressure and gesture experiments have produced promising results, but the evidence still comes from small tests involving healthy volunteers.
Nine new cases expand the picture of this ultra-rare metabolic disease from encephalopathy to cardiac and fetal abnormalities; four children treated with niacin survived subsequent fevers or infections, but the limited case experience is not enough to demonstrate efficacy.
A two-year Phase 3 trial showed that the oral complement factor B inhibitor iptacopan can slow kidney function decline in high-risk adults, converting the accelerated approval originally based on a surrogate endpoint into full approval; however, the risks of serious infections, vaccination requirements, and REMS management remain treatment barriers.
A large-scale repository spanning 25 cancer types links tumor models with multi-omics data and clinical information. Most models faithfully preserve the characteristics of the original tumors, but changes in cellular states caused by culture conditions also define their limitations for researchers.
HoneyNaps’ software can detect respiratory events from multichannel sleep-study signals and distinguish obstructive, central, and mixed apnea; however, 510(k) clearance and the high agreement rates reported by the company cannot yet replace validation of benefits in real-world clinical settings.
One independent comparison found that clinical AI built specifically for physicians does not necessarily outperform frontier general-purpose models; another physician evaluation placed OpenEvidence far ahead. The divergence concerns more than rankings, exposing how data sources, scoring methods, and conflicts of interest shape the definition of “reliable.”
Research has found that phosphorylated tau can disrupt mitochondrial electron flow, creating a vicious cycle of oxidative stress and further tau abnormalities. Blocking the mechanism provided protection in multiple models, but the evidence remains at the preclinical stage.
MonumenTAL-6 is the first randomized Phase 3 trial to simultaneously target BCMA and GPRC5D; interim results show a survival benefit, but complete efficacy, safety, and treatment-discontinuation data have yet to be disclosed.
The combination of Tudriqev and nivolumab has finally cleared the threshold for market entry, but the approval is still based on a controversial single-arm trial; its true clinical benefit must await answers from a randomized trial whose primary analysis is not expected to be completed until 2029.