The Evidence Layer for Scientific AI

Physics & fundamental questions

Wavefunction: mathematics versus physical reality

Quantum theory uses the wavefunction to calculate probabilities, while experiments do not yet select one physical interpretation; competing interpretations can share predictions.

Dark energy: measured acceleration, unknown cause

Accelerated cosmic expansion is well measured, but the physical source remains unidentified and evolving-dark-energy claims require stronger cross-survey confirmation.

Biological oscillators and information capacity

A quality-factor screen can reject frequency-carrier proposals as necessary conditions, but conclusions depend on estimated coherence, persistence, and readout times.

Extremophile adaptation and physical constraints

Genome-wide motif patterns may track temperature-linked DNA properties, although current evidence is phylogeny-adjusted correlation rather than direct causal proof.

Life, health & disease

Bee microbiomes clear a plant toxin intermediate

In vivo metabolite patterns, strain assays, proteomics, and GH3 validation support microbial prunasin clearance, while final hydrogen-cyanide measurements remain incomplete.

Estrogen receptor α controls mouse puberty timing and completion

Conditional deletion in Kiss1 neurons advances puberty onset but disrupts later ovulation, supporting nucleus-specific roles in mice with limited human extrapolation.

Visual deprivation blocks intrinsic cortical gain

Electrophysiology, Kv2.1 localization, pharmacology, and reversibility support failed intrinsic plasticity in deprived L5 neurons, though findings are layer- and species-limited.

NUDT5 links thiopurine metabolism to leukemia response

CRISPR loss and rescue connect NUDT5 to thiopurine nucleotide pools and resistance in patient-derived leukemia models, requiring broader clinical validation.

Immunometabolism as a context-dependent control system

A narrative review links nutrient competition, metabolite signaling, and epigenetic regulation to immune state, but provides no new dataset or quantitative synthesis.

Genomics, evolution & ecology

Starship transposons and fungal virulence evolution

Comparative genomics links Starship regions to genome plasticity, virulence cargo, and NLP6 function, while broad causal claims remain association-heavy.

Quinoa metabolite genetics across tissues and drought

Multi-tissue mGWAS and functional tests identify candidate biosynthetic genes, while drought network links remain mainly correlational.

Multiple origins of Trypanosoma evansi

Microsatellite clustering supports at least two Kenyan T. brucei backgrounds, but genome-wide markers are needed to resolve origin count and mechanism.

Long reads improve antimicrobial-resistance linkage

Long-read and hybrid assemblies can connect resistance genes to plasmids and hosts, although mobility must be validated rather than assumed from linkage alone.

Spatial competition and resident diversity

Localized resource use may suppress invaders while preserving distant niches, but direct evidence for diversity protection remains a testable ecological inference.

Tree microhabitats structure micro-food webs

Community surveys and stable isotopes reveal distinct microhabitat assemblages and trophic niches, with seasonal and single-site limits on generality.

Cells, brains & development

NF-κB regulation at accessible and closed chromatin

Cross-species chromatin profiling and CCL2 perturbations support conserved NF-κB binding modes, while genome-wide causality beyond tested loci remains uncertain.

Partial CD45 exclusion during T-cell signaling

Live-cell imaging supports partial phosphatase exclusion at signaling contacts, but discrimination causality still relies partly on modeling rather than direct fraction manipulation.

Thymic progenitor for KIR-positive NK cells

Flow analysis, single-cell data, differentiation assays, and thymic-defect comparisons support a thyILC1 population, without direct human in vivo lineage tracing.

Primate dorsal cochlear nucleus synaptic architecture

Electron microscopy and receptor localization support conserved parallel-fiber synapses in rhesus tissue, while functional computation remains untested.

FGF4, ERK, and inner-cell-mass fate prediction

Bistable and tristable models support FGF–ERK threshold logic, but DyMoTree pseudotime adds an untested predictive claim after Nanog and Gata6 adjustment.

Cortical GABA dynamics during motor learning

7T MRS finds learning-specific GABA reduction and early-state associations with later performance, but metabolite measures cannot establish synaptic causality.

Environment, organisms & systems

Biofilms accelerate arsenite oxidation in filters

Time-dependent kinetics and inhibition support a shift toward biological oxidation, although manganese mass balance and inhibitor specificity limit certainty.

Sea-slug kleptoplast longevity depends on algal donor

Donor identity strongly predicts photosynthetic persistence during starvation, while molecular repair and photoprotection mechanisms remain unresolved.

Ectomycorrhiza changes pine nitrogen and water economy

Axenic experiments show biomass and allocation changes without guaranteed photosynthetic stimulation, but causal nitrogen accounting is incomplete.

Drosophila microbiomes tune circadian transcription

Sterile and colonized comparisons support altered rhythmic amplitude and stability under stress, with simplified microbiomes limiting ecological scope.

Social context reshapes naked mole-rat oxygen choice

Choice assays show social pairing shifts preferred hypoxia levels, while proposed energy-demand mechanisms lack concurrent physiological measurements.

The Evidence Layer for Scientific AI

Ask questions across claims linked to experiments, exact results, limitations, and sources.Know what the science actually supports before you trust the answer.

Explore by Goal

Physics & fundamental questions

Life, health & disease

Genomics, evolution & ecology

Cells, brains & development

Environment, organisms & systems

Methods, technology & evidence checks

Ask a Follow-Up

Stay current without chasing every paper.

My BGPT

Trending