Base by Base

Yan X et al., The American Journal of Human Genetics, Corrected proof. doi:10.1016/j.ajhg.2025.12.008 - In 569 TwinsUK subcutaneous adipose biopsies, twin models and GWAS identify SCD, FADS and 3p25.2 (PPARG) loci regulating fatty acid levels and conversions. Key terms: SCD, FADS1, PPARG, adipose-tissue, fatty-acids.

Study Highlights:
System and sample: 569 female TwinsUK subcutaneous adipose biopsies with matched serum, RNA-seq and 450K methylation data were analyzed. Key methods: gas chromatography fatty acid profiling, ACE twin heritability models, GWAS of 18 fatty acids and 15 product-to-precursor ratios, colocalization with adipose eQTLs and meQTLs, and polygenic score analysis. Main quantitative results: heritability of individual fatty acids ranged from 5%–59% while 15 fatty acid ratios were heritable and GWAS identified 10 genome-wide significant loci including SCD and FADS with the SCD lead variant explaining ~7–11% of variance and ratios showing heritability up to ~54%. Functional implication: colocalizations with adipose-specific eQTLs and meQTLs and associations between metabolic polygenic scores and fatty acid levels link local genetic regulation in adipose tissue to renal and cardio-metabolic phenotypes.

Conclusion:
Local genetic variation in adipose tissue, including SCD, FADS and a 3p25.2 locus near PPARG, regulates fatty acid composition and conversion and is connected to renal and cardio-metabolic traits.

Music:
Enjoy the music based on this article at the end of the episode.

Article title:
Genetic regulation of fatty acid content in adipose tissue

First author:
Yan X

Journal:
The American Journal of Human Genetics, Corrected proof. doi:10.1016/j.ajhg.2025.12.008

DOI:
10.1016/j.ajhg.2025.12.008

Reference:
Yan X., Roberts A.L., El-Sayed Moustafa J.S., Villicaña S., Al-Hilal M., Tomlinson M., Menni C., Sanders T.A.B., Freidin M.B., Bell J.T., Small K.S., Genetic regulation of fatty acid content in adipose tissue. The American Journal of Human Genetics 113, 1–18, February 5, 2026. https://doi.org/10.1016/j.ajhg.2025.12.008

License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) - https://creativecommons.org/licenses/by/4.0/

Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00

Official website https://basebybase.com

On PaperCast Base by Base you’ll discover the latest in genomics, functional genomics, structural genomics, and proteomics.

Episode link: https://basebybase.com/episodes/scd-fads-pparg-adipose

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-02-04.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited substantive scientific claims regarding adipose fatty acid heritability and tissue specificity, GWAS loci (SCD, FADS, 3p25.2/PPARG), adipose molecular mechanisms (eQTL/meQTL colocalization), the fat–kidney link via pleiotropy, and polygenic-score associations with adipose fatty acids; also covered study limitat
- transcript topics: Adipose fatty acid heritability and tissue specificity; Genome-wide association signals for adipose fatty acids (SCD, FADS, 3p25.2/PPARG); Adipose-specific regulation and colocalization with eQTL/meQTL; Arachidonic acid and kidney trait pleiotropy; Polygenic scores and adipose fatty acids; Limitations and demographic scope

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 8
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- Adipose fatty acids show heritability with tissue specificity; adipose heritability ranges widely (0–59%), with some fatty acids showing little to no heritability in adipose but hi
- GWAS identified 10 independent loci associated with adipose fatty acids and ratios; SCD and FADS loci were replicated, and a novel 3p25.2 locus near PPARG/MKRN2/TSEN2 was identifie
- SCD locus exhibits adipose-specific regulatory effects (adipose eQTL and meQTL colocalization; adipose SCD expression correlates with fatty acid ratios).
- 3p25.2 locus regulates arachidonic acid and the arachidonic acid/linoleic acid ratio in adipose tissue and colocalizes with kidney-trait signals (eGFR, creatinine, cystatin C), ind
- GWAS signals at SCD, FADS, and 3p25.2 show molecular-trait colocalization with eQTLs/meQTLs, supporting tissue-specific regulatory mechanisms.
- Polygenic scores for BMI, fat distribution, and triglycerides are associated with adipose tissue fatty acid levels, linking systemic metabolic risk to local adipose composition.

QC result: Pass.

Guest O et al. - Position piece urging universities to resist uncritical adoption of AI technologies such as LLMs and chatbots because they undermine academic freedom, integrity, and pedagogical skills. Key terms: artificial intelligence, higher education, large language models, academic freedom, critical AI literacy.

Study Highlights:
System: the higher education sector and university classrooms; methods: a co-authored open letter, conceptual analysis, and literature synthesis drawing on historical and contemporary sources. The authors analyse how AI industry marketing, ambiguous jargon, closed-source models, and extractive data and labour practices create institutional dependencies and conflicts of interest. They show this structural entanglement erodes research integrity, deskills students and staff, and produces environmental and social harms. As a functional implication they call for principled refusal, transparency, critical AI literacy, and policy measures to protect academic freedom and the ecosystem of human knowledge.

Conclusion:
Universities must reject the uncritical adoption of AI technologies and take active measures to safeguard critical thinking, expertise, academic freedom, and scientific integrity.

Music:
Enjoy the music based on this article at the end of the episode.

Article title:
Against the Uncritical Adoption of AI Technologies in Academia

First author:
Guest O

Journal:
Zenodo

DOI:
10.5281/zenodo.17065099

Reference:
Guest O, Suarez M, Muller BCN, van Meerkerk E, Oude Groote Beverborg A, de Haan R, et al. Against the Uncritical Adoption of AI Technologies in Academia. Zenodo. 2025. https://doi.org/10.5281/zenodo.17065099

License:
This episode is based on an open-access preprint released under the Creative Commons Attribution 4.0 International License (CC BY 4.0) - https://creativecommons.org/licenses/by/4.0/

Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00

Official website https://basebybase.com

On PaperCast Base by Base you’ll discover the latest in genomics, functional genomics, structural genomics, and proteomics.

Episode link: https://basebybase.com/episodes/universities-resist-ai-adoption

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-02-02.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited transcript sections that discuss the preprint 'Against the Uncritical Adoption of AI Technologies in Academia' and its critiques: hype/terminology, environmental and labor harms, two victims of plagiarism, knowledge production vs knowledge translation, and guardrails for AI-assisted science communication.
- transcript topics: AI hype and terminology in academia; Marketing, hype, and harm of AI technologies; Environmental costs and ghost labor in AI; Two victims of plagiarism (original author and audience); Knowledge production vs knowledge translation; Guardrails for ethical AI use in science communication

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 7
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- AI hype and marketing as a concern in academia
- Environmental costs and labor (ghost work) associated with AI systems
- The concept of two victims of plagiarism (original author and audience)
- Distinction between knowledge production and knowledge translation
- Need for guardrails and human verification in AI-assisted science communication
- Call for critical AI literacy and public discussion

QC result: Pass.

Rashid AI et al., The American Journal of Human Genetics, Corrected proof. doi:10.1016/j.ajhg.2025.12.012 - US antitrust shifts in human genetic technologies: FTC scrutiny of Illumina’s acquisition of Grail alters NGS market oversight and could affect spin-offs and startups. Key terms: Illumina, Grail, vertical mergers, FTC merger guidelines, non-compete rule.

Study Highlights:
This perspective examines the US legal and regulatory landscape for human genetic and genomic technologies, focusing on FTC and DOJ policy changes and enforcement habits. Using a case-study approach centered on Illumina’s proposed acquisition and eventual divestiture of Grail, the authors review updated merger guidelines, non-compete rule developments, and exclusive-dealing concerns. They report that the 2023 merger guidelines lowered thresholds for presumed anti-competitiveness and that the FTC’s challenge emphasized risks to nascent mult-cancer early detection competitors that rely on next-generation sequencing platforms. The authors conclude that heightened vertical-merger scrutiny may reduce spin-offs and early-stage acquisitions, prompting firms to favor in-house R&D or alternate collaboration models.

Conclusion:
Heightened scrutiny of vertical mergers and attention to nascent competition in US antitrust policy is likely to reshape biotechnology strategies by reducing spin-offs and encouraging in-house development.

Music:
Enjoy the music based on this article at the end of the episode.

Article title:
Competition in human genetic technologies: The current US legal landscape

First author:
Rashid AI

Journal:
The American Journal of Human Genetics, Corrected proof. doi:10.1016/j.ajhg.2025.12.012

DOI:
10.1016/j.ajhg.2025.12.012

Reference:
Rashid AI, Rincon NA, Rihani N, Wagner JK. Competition in human genetic technologies: The current US legal landscape. The American Journal of Human Genetics. 2026;113:1–10. https://doi.org/10.1016/j.ajhg.2025.12.012

License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) - https://creativecommons.org/licenses/by/4.0/

Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00

Official website https://basebybase.com

On PaperCast Base by Base you’ll discover the latest in genomics, functional genomics, structural genomics, and proteomics.

Episode link: https://basebybase.com/episodes/illumina-grail-vertical-mergers

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-02-02.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited the transcript sections that discuss antitrust policy landscape, the Illumina/Grail vertical merger case, regulatory instruments (HSR, Sherman, Clayton), NASCENT competition, spin-offs and IRS spin-off rules, non-compete status, and startup financing implications.
- transcript topics: Illumina/Grail vertical merger case; Foreclosure theory in vertical mergers; Open offer and divestiture outcome; 2023 merger guidelines and vertical mergers; Nascent competition and spin-offs; IRS rules on corporate spin-offs

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 8
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- Illumina/Grail case as a vertical merger example with divestiture
- Foreclosure as a potential anti-competitive mechanism in vertical mergers
- 2023 merger guidelines lowered thresholds for presumed anti-competitiveness
- IRS spin-off rules tightened in January 2025 with stricter proof and reporting
- Non-compete regulation status: initial nationwide ban with subsequent rulings and continued case-by-case enforcement
- Startup financing pressures (kill zone) due to antitrust scrutiny and exit constraints

QC result: Pass.

Morsy H et al., The American Journal of Human Genetics, Corrected proof. doi:10.1016/j.ajhg.2025.12.015 - Exome sequencing identifies homozygous MDGA2 loss-of-function variants in nine individuals and functional neuronal assays show impaired MDGA2 trafficking with disrupted Nlgn1-dependent excitatory synapse regulation causing DEE. Key terms: MDGA2, developmental and epileptic encephalopathy, loss-of-function, neuroligin-1, exome sequencing.

Study Highlights:
Exome sequencing of consanguineous families identified seven distinct homozygous MDGA2 loss-of-function variants in nine individuals with severe developmental and epileptic encephalopathy. Functional evaluation used mammalian expression in HEK293T cells, heterologous synapse-formation assays, cultured hippocampal neurons, and electrophysiology. Representative nonsense variants abolished MDGA2 surface trafficking, disrupted MDGA2–Nlgn1 binding, failed to suppress excitatory synapse density, and did not reduce AMPAR- and NMDAR-mediated synaptic responses. These synaptic deficits imply disruption of excitatory-inhibitory balance, providing a mechanistic link to early-onset intractable seizures and progressive neurodevelopmental impairment.

Conclusion:
Homozygous MDGA2 loss-of-function variants cause an autosomal-recessive developmental and epileptic encephalopathy by impairing MDGA2 trafficking and Nlgn1-dependent suppression of excitatory synapses.

Music:
Enjoy the music based on this article at the end of the episode.

Article title:
MDGA2 homozygous loss-of-function variants cause developmental and epileptic encephalopathy

First author:
Morsy H

Journal:
The American Journal of Human Genetics, Corrected proof. doi:10.1016/j.ajhg.2025.12.015

DOI:
10.1016/j.ajhg.2025.12.015

Reference:
Morsy H, Kim H, Jang G, et al. MDGA2 homozygous loss-of-function variants cause developmental and epileptic encephalopathy. The American Journal of Human Genetics. 2026;113:1–12. https://doi.org/10.1016/j.ajhg.2025.12.015

License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) - https://creativecommons.org/licenses/by/4.0/

Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00

Official website https://basebybase.com

On PaperCast Base by Base you’ll discover the latest in genomics, functional genomics, structural genomics, and proteomics.

Episode link: https://basebybase.com/episodes/mdga2-loss-of-function-dee

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-02-01.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audit covered the sections describing MDGA2 function in synaptic regulation, genetic basis (LoF variants in consanguineous families), functional assays (trafficking, Nlgn1 interactions, synapse formation, electrophysiology), clinical phenotype (DEE with seizures), and therapeutic angles (ketogenic diet, potential recep
- transcript topics: MDGA2 as a brake on Nlgn1 and suppression of excitatory synapses; Genetic basis: homozygous loss-of-function MDGA2 variants in consanguineous families; Functional validation: trafficking and surface expression of MDGA2; Nlgn1 interaction; Synapse formation and electrophysiology: mEPSCs, AMPAR/NMDAR EPSCs; Clinical presentation: DEE phenotype, hypotonia, seizures, dysmorphic features, MRI findings; Therapeutic angles: ketogenic diet effects; potential TRKB/AMPA receptor–related targets

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 6
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- MDGA2 homozygous loss-of-function variants identified in nine individuals from seven consanguineous families
- MDGA2 variants cause trafficking impairment and loss of surface expression with disrupted Nlgn1 interaction
- MDGA2 waives suppression of excitatory synapses in LoF variants; WT MDGA2 suppresses excitatory synapse numbers
- Electrophysiology shows MDGA2 WT reduces mEPSC frequency and AMPAR/NMDAR EPSCs; LoF variants fail to alter synaptic function
- DEE phenotype with early-onset seizures, severe developmental delay, dysmorphic features; brain MRI shows delayed myelination and early brain atrophy
- Ketogenic diet produced partial seizure control in two individuals (P4 and P5B)

QC result: Pass.

Avsec et al., Nature, doi:10.1038/s41586-025-10014-0 - AlphaGenome, a 1 Mb DNA deep‑learning model, predicts base‑pair‑resolution genome tracks (RNA‑seq, splicing, chromatin) and scores variant effects, achieving state‑of‑the‑art performance across modalities. Key terms: AlphaGenome, splicing, eQTL, chromatin-accessibility, 1Mb-sequence.

Study Highlights:
AlphaGenome is a unified sequence‑to‑function deep learning model trained on human and mouse genomes that consumes 1 Mb of DNA and predicts 5,930 human genome tracks across 11 modalities using a U‑Net‑inspired encoder, transformer tower and decoder. The model was pretrained with fold splits and distilled into a single student model for efficient variant scoring, enabling base‑pair resolution outputs and splice junction prediction alongside splice site usage and RNA‑seq coverage. Quantitatively, AlphaGenome outperformed or matched external models on 22 of 24 genome track tasks and on 25 of 26 variant effect benchmarks, improving eQTL sign prediction and QTL effect correlations. The multimodal outputs enable mechanistic interpretation of variants, for example recapitulating oncogenic TAL1 enhancer mutations and identifying splice‑disrupting variants.

Conclusion:
AlphaGenome provides a unified 1‑Mb multimodal, base‑resolution sequence model that substantially improves genome track and regulatory variant effect prediction and enables mechanistic, cross‑modality interpretation.

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-01-30.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited the transcript portions presenting AlphaGenome architecture, multimodal outputs, benchmarking, splicing, TAL1 mechanism, 3D genome, in silico mutagenesis, and limitations.
- transcript topics: AlphaGenome architecture (1 Mb input, U‑Net backbone, transformers, sequence parallelism); Multimodal genome tracks (5,930 human tracks, 1,128 mouse tracks across 11 modalities); Variant effect benchmarking (26 tasks; 25/26 ahead); Splicing variant predictions (splice sites, splice junctions, usage); TAL1 oncogene mechanism (neo-enhancer, MYB motif); 3D genome contact maps predictions

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 8
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- AlphaGenome takes input of 1 Mb DNA and predicts thousands of genome tracks across 11 modalities at base-pair resolution
- 5,930 human genome tracks and 1,128 mouse genome tracks predicted
- AlphaGenome matched or outperformed strongest external models on 25 of 26 variant-effect benchmarks
- Splicing variant predictions include splice sites, splice junctions, and splice site usage
- TAL1 oncogene mechanism recapitulated via a neo-enhancer and MYB motif
- 3D genome folding predicted via contact maps with improved performance over specialized tools

QC result: Pass.

Sierant MC et al., Proc. Natl. Acad. Sci. U.S.A. 2025.122:e2420343122 - MIPseq and exome sequencing of 11,555 human congenital heart disease probands implicate 60 dominant CHD genes, with NOTCH1 cysteine‑altering and transmitted MYH6 missense variants driving distinct defects. Key terms: congenital heart disease, NOTCH1, MYH6, MIPseq, de novo mutations.

Study Highlights:
We analyzed 11,555 human CHD probands from PCGC and PHN using a 248‑gene MIPseq panel and whole‑exome sequencing. A meta‑analysis of de novo and very rare transmitted/unphased damaging variants identified 60 genes with significant burden, accounting for damaging variants in 10.1% of probands with similar DNM and transmitted contributions. Mechanistically, NOTCH1 missense mutations that introduce or remove cysteines in EGF domains were highly enriched in tetralogy of Fallot and conotruncal defects, while transmitted damaging MYH6 missense variants were overtransmitted and contributed to multiple CHD subtypes. Genes with cardiomyocyte‑restricted expression correlated with isolated CHD, whereas broadly brain‑expressed genes correlated with neurodevelopmental delay, supporting genotype‑informed risk assessment.

Conclusion:
Targeted genomic analysis of 11,555 CHD probands identifies 60 dominant genes accounting for 10.1% of cases and supports molecular diagnosis to stratify cardiac and neurodevelopmental risk.

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-01-30.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited the transcript sections describing the study design (MIPseq), cohort composition, main genetic findings (60 genes, 10.1%), NOTCH1 cysteine-altering variants, MYH6 transmission findings, chromatin gene involvement, syndromic diagnoses, and cost reductions and future gene discovery projections.
- transcript topics: MIPseq methodology and cost reduction; Cohort assembly and the 248-gene panel; 60 significant CHD genes and 10.1% probands; NOTCH1 cysteine-altering variants in TOF/CTD (EGF domains, domain #5); MYH6 transmitted damaging variants and penetrance; Chromatin modifier genes and neurodevelopmental/EC phenotypes

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 6
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- Study analyzed 11,555 CHD probands with a 248-gene MIPseq panel.
- 60 genes identified with significant burden of damaging variants; 10.1% of probands affected.
- Equal contributions from de novo and transmitted variants to CHD risk within the 60-gene set.
- NOTCH1 missense variants introducing/removing cysteines in EGF domains are enriched in TOF and CTD; cysteine-altering variants cluster in EGF domain #5.
- Transmitted damaging MYH6 missense variants enriched across multiple CHD phenotypes; risk ratio ~6; 2–3% of probands; ~1% of all trios.
- MYH6 is heart-restricted in expression; NDD risk among MYH6 carriers is low (~4%).

QC result: Pass.

PNAS - RPE-specific MCT2 gene delivery preserves cones and vision in retinitis pigmentosa models

Music:
Enjoy the music based on this article at the end of the episode.

Article title:
RPE-specific MCT2 expression promotes cone survival in models of retinitis pigmentosa

Journal:
PNAS

DOI:
10.1073/pnas.2421978122

License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) – https://creativecommons.org/licenses/by/4.0/

Support:
Base by Base – Stripe donations: https://donate.stripe.com/7sY4gz71B2sN3RWac5gEg00

Official website https://basebybase.com

On PaperCast Base by Base you’ll discover the latest in genomics, functional genomics, structural genomics, and proteomics.

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-01-29.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited transcript sections covering: metabolic mechanism (RPE lactate uptake via MCT2 and glucose sparing), vector design (AAV8.Best1.MCT2), RP animal models and cone survival/function outcomes, FLIM-based lactate/glucose sensors (LiLac, GlucoSnFR-TS), MCT2 localization to apical/basal RPE membranes, observed species-
- transcript topics: RPE-specific MCT2 metabolic reprogramming; Retina metabolism: lactate, glucose, and NAD+ dynamics in RP; AAV8.Best1.MCT2 vector design and RPE specificity; RP models: S334ter rat, FVB mouse, P23H mouse; Cone survival and functional outcomes (cone counts and optomotor assay); FLIM-based lactate and glucose sensors (LiLac, GlucoSnFR-TS)

QC Summary:
- factual score: 10/10
- metadata score: 8/10
- supported core claims: 6
- claims flagged for review: 0
- metadata checks passed: 3
- metadata issues found: 1

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- RPE-specific MCT2 expression increases cone survival across RP models (rat S334ter; mouse FVB and P23H).
- MCT2 expression transiently preserves cone function in P23H mice (optomotor results at P40; diminished by P53).
- MCT2 localizes to apical and basal membranes of the RPE after AAV delivery.
- FLIM sensors show higher intracellular lactate and greater glucose accumulation in MCT2-expressing RPE, consistent with reduced glycolysis and glucose sparing for cones.
- Three RP mutations/models tested support a gene-agnostic approach.
- Toxicity observed in mice but not rats; potential translational bottleneck due to species-specific responses.

QC Flagged Items (audited and not fully supported):
- license: metadata does not match the canonical article record. License naming differs between metadata snapshot and article description; both refer to CC BY variants.
Internal QC note: manual editorial review is recommended before publication.

QC result: Warning. Items above were flagged during automated QC; the editorial team reviewed them before release.

Gori K et al., Proc. Natl. Acad. Sci. U.S.A. 2025.122:e2424634122 - In canine transmissible venereal tumor (CTVT), deep sequencing and cytogenetics identify a 15‑Mb horizontally transferred nuclear element (N-HT1) acquired ~2,000 years ago that is transcriptionally active. Key terms: CTVT, horizontal gene transfer, N-HT1, PacBio long-read sequencing, centromeric fusion.

Study Highlights:
The authors screened 174 transmissible tumor genomes, focusing on CTVT, DFT1, and DFT2, using deep short-read sequencing, long-read PacBio sequencing, structural variant analysis, and metaphase FISH. In CTVT-A they discovered a 15-Mb dicentric element (N-HT1) assembled from 11 fragments of six chromosomes that forms the short arm of a small submetacentric chromosome after centromeric fusion. Mutation density and CpG-based dating place N-HT1 acquisition about 2,000 years ago, and transcriptome allele deconvolution shows N-HT1 is transcriptionally active and adopts the CTVT expression profile. Functional interrogation found no clear oncogenic drivers on N-HT1, with at least one rescued gene (ARFGEF3) later inactivated, consistent with the element behaving as a likely neutral passenger.

Conclusion:
A single host-to-tumor nuclear horizontal transfer event was detected in sampled transmissible cancers: CTVT-A acquired a 15-Mb N-HT1 element that is transcriptionally active but shows no clear evidence of positive selection.

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-01-28.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited the core scientific claims presented about N-HT1: its structure, origin, timing, transcriptional activity, and evolutionary implications, as described in the article.
- transcript topics: CTVT background and transmissible cancers; SNP flipping screen to detect host-to-tumor nuclear transfer; Discovery and structure of N-HT1 (15 Mb, 11 fragments, from 6 chromosomes); Cytogenetics and centromeric fusion of N-HT1 onto a CTVT chromosome; Timing: acquisition ~2000 years ago and donor ancestry; Gene expression and functional impact (ARFGEF3 rescue)

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 8
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- N-HT1 is a 15-Mb horizontally transferred nuclear DNA element composed of 11 fragments from six canine chromosomes
- N-HT1 forms the short arm of a small submetacentric chromosome via centromeric fusion onto a CTVT chromosome
- N-HT1 was acquired by CTVT-A approximately 2,000 years ago
- N-HT1 is transcriptionally active and adopts the CTVT expression profile
- There is no clear evidence of positive selection acting on N-HT1 within CTVT-A
- ARFGEF3 was rescued from a null state by N-HT1 but was subsequently inactivated again

QC result: Pass.

Ju X-C et al., Proc. Natl. Acad. Sci. U.S.A. 2025.122:e2508540122 - Human-specific ADSL A429V substitution and a common regulatory haplotype reduce ADSL activity and raise purine substrates in the brain, altering mouse behavior. Key terms: adenylosuccinate lyase, A429V, purine biosynthesis, succinyladenosine, human evolution.

Study Highlights:
Model: mice humanized for ADSL carrying the modern-human A429V (with R428Q) were compared to wild-type littermates using ultraperformance LC–Orbitrap metabolomics and automated IntelliCage behavioral assays. Mechanistic/quantitative result: SAICAr and S-Ado concentrations increased up to ~2-fold in liver and 1.8–5.4-fold across cerebrum regions, and these increases correlated negatively with Adsl mRNA expression across tissues. Human genetics: a 7.8-kb haplotype (including rs8135371) at >97% carrier frequency is associated with lower ADSL expression, higher S-Ado in cerebrospinal fluid, and signals of positive selection. Functional implication: female humanized mice accessed water more efficiently under restricted conditions, linking reduced ADSL activity to altered behavior.

Conclusion:
Two genetic changes on the modern human lineage—a nearly fixed A429V amino acid substitution and a common regulatory haplotype—have reduced ADSL activity and expression, increasing purine substrates particularly in the brain and producing measurable behavioral effects in mice.

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-01-27.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited transcript sections describing the ADSL enzyme, the A429V substitution, mouse model (humanized Adsl), tissue metabolite changes, the IntelliCage water-competition experiment with female-specific effects, regulatory haplotype and selection signals, and human CSF/metabolite correlations with intelligence.
- transcript topics: ADSL role in purine biosynthesis; A429V modern-human substitution and enzyme stability; CRISPR-based humanized Adsl mouse model; Tissue-specific metabolic substrates SAICAr and S-Ado; Brain- and liver-specific accumulation patterns; IntelliCage water-competition behavioral assay and female-specific effects

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 6
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- The modern-human ADSL coding change (A429V) reduces enzyme stability and activity relative to Neanderthal version.
- A second regulatory haplotype reduces ADSL expression, especially in brain, with evidence of positive selection.
- In humanized Adsl mice, SAICAr and S-Ado accumulate, particularly in brain and liver; brain shows greatest impact due to low baseline Adsl expression.
- Female humanized mice show increased water-approach behavior under restricted water conditions; males show no difference.
- CSF levels of S-Ado correlate negatively with intelligence in humans (rg ≈ -0.13; small effect size but statistically significant).
- Positive selection signals are present for both the coding and regulatory changes, with estimated selection coefficients around 0.0016–0.0018.

QC result: Pass.

Morris TT et al., Proc. Natl. Acad. Sci. U.S.A. 2026.123:e2516460123 - EA and cognition polygenic indexes (PGIs) in three British birth cohorts show EA PGI associations with years of education increased from 1946–1970 and were strongest in advantaged socioeconomic backgrounds. Key terms: educational attainment, polygenic index, gene-environment interaction, British birth cohorts, socioeconomic status.

Study Highlights:
Using three nationally representative British birth cohorts born 1946, 1958, and 1970, the authors analyzed polygenic indexes for educational attainment (EA) and cognition. They generated PGIs with clumping-and-thresholding (PRSice2) and LDpred2, used multiple imputation and inverse probability weighting, and estimated linear models including cohort-by-PGI interactions. EA PGI associations increased from approximately 0.44 to 0.67 years of education per 1-SD and incremental R2 rose from 3.5% to 5.1% across cohorts, while cognition PGI associations were broadly stable. There was strong evidence of gene–environment interaction: returns to EA genetic liability were disproportionately larger among those born into more advantaged socioeconomic backgrounds.

Conclusion:
Across three British birth cohorts born 1946–1970, genetic liability indexed by an EA PGI became more predictive of years of completed education while cognition PGI prediction remained stable, and EA PGI effects were amplified in advantaged socioeconomic contexts.

QC:
This episode was checked against the original article PDF and publication metadata for the episode release published on 2026-01-25.

QC Scope:
- article metadata and core scientific claims from the narration
- excludes analogies, intro/outro, and music
- transcript coverage: Audited the core scientific claims conveyed in the transcript: temporal increase of EA PGI associations, stability of cognition PGI, GxE interaction by SES, non-cognitive vs cognitive PGI interpretation, and study limitations.
- transcript topics: EA PGI prediction of education across cohorts; Cognition PGI stability across cohorts; Polygenic indices (EA PGI and cognition PGI) methods; Gene–environment interactions with socioeconomic background; Non-cognitive vs cognitive PGI interpretation; Limitations: ancestry and PGI portability

QC Summary:
- factual score: 10/10
- metadata score: 10/10
- supported core claims: 5
- claims flagged for review: 0
- metadata checks passed: 4
- metadata issues found: 0

Metadata Audited:
- article_doi
- article_title
- article_journal
- license

Factual Items Audited:
- EA PGI associations with years of completed education increased across cohorts (1946c 0.44, 1958c 0.49, 1970c 0.67).
- Cognition PGI associations with education remained broadly stable across cohorts (1946c 0.23, 1958c 0.27, 1970c 0.24).
- EA PGI explained incremental variance in education rising from 3.5% to 5.1% across cohorts.
- Gene–environment interaction: higher returns to EA PGI among advantaged backgrounds and near-null returns among the most disadvantaged.
- Non-cognitive PGI captures additional variance; cognition PGI did not show increasing predictive power.
- Limitations include ancestry limitation to white European participants and PGI portability concerns.

QC result: Pass.