Longevity, genes,
and disease-modifying AI.

GRAIL Galleri MCED — FDA Molecular & Clinical Genetics Panel meeting today (2026-09-23)

Prescription methylation + NGS multi-cancer early detection blood test PMA under AdComm review (adults ≥50; Cancer Signal Origin prediction); hybrid meeting 9:00 a.m.–6:00 p.m. ET; panel votes on reasonable assurance of safety/effectiveness and benefit–risk (nonbinding; FDA decides).

Caveat: Advisory committee / pre-PMA decision — not an approval; adjunct to (not replacement for) guideline single-cancer screening; verify against FDA briefing materials and final vote when posted.

CRISPR-Cas12a corrects cryptic PCCA pseudoexon in propionic acidemia cell models (CSIC/UAM · Molecular Therapy Nucleic Acids)

Deep/cryptic splice mutation (hidden outside coding sequence) disrupted enzyme production; CRISPR-Cas12a targeting the PCCA pseudoexon / splice-enhancer restored correct splicing and recovered PCCA/PCCB protein / PCC activity (~30% of normal in best guide reports) in human cell models of this rare metabolic disease.

Caveat: Experimental cell proof-of-concept — not a patient therapy; no clinical delivery/safety data yet; diagnostic insight for mutations standard tests can miss.

Ex vivo base editing of Artemis (DCLRE1C) mutations for ART-SCID (Advanced Biotechnology)

Cytidine/adenine base editors corrected three pathogenic Artemis variants in cell models (~50% CBE at c.181T>C; ~35% ABE at c.49G>A; ~20% ABE at c.404G>A) with no detectable off-targets at predicted sites; partial restoration of Artemis endonuclease activity for some alleles — path toward gene therapy where transplant is hard in radiosensitive SCID.

Caveat: 293T / reporter models only — not patient HSPCs yet; one allele failed functionally due to damaging bystander (D136G); genome-wide off-target and engraftment work still needed.

USP22–SPI1–NAMPT axis protects bone-marrow MSCs from oxidative senescence (Molecular Genetics and Genomics)

Deubiquitinase USP22 stabilizes SPI1, which transcriptionally activates NAMPT → NAD+ salvage; overexpression blunted H₂O₂-induced senescence and preserved osteogenic differentiation in BMSCs; SPI1 also down in OVX-rat osteoporosis marrow MSCs.

Caveat: Preclinical (cells + rat) — USP22/SPI1 have oncogenic contexts elsewhere; not a human anti-aging therapy; translational risk/benefit open.

49 tissue-specific transcriptomic aging clocks map mixed drug “age” scores across tissues (Biogerontology)

GTEx-based elastic-net clocks per tissue (median r≈0.54; aorta highest); LINCS L1000 projection → vast drug×tissue age-modulatory matrix; most compounds show mixed directions across tissues — AI/computational geroscience for clinicians, not a single “anti-aging drug” list.

Caveat: Hypothesis-generating / transcriptional endpoints ≠ lifespan; after FDR only a small fraction of pairs significant; rapamycin showed net pro-aging transcript scores here — author stresses that does not overturn animal lifespan evidence.

Research AI ≠ clinical cure claims

Pre-PMA advisory votes, cell-model gene editing, preclinical senescence axes, and transcriptomic drug×tissue clocks are different evidence classes. This site keeps them labeled so a headline never collapses into “AI cured aging.”