June 2025 — Monthly Briefing

1. What was read

832 research articles from 32 of the 33 journals on the roster.

A further 277 records (25% of the 1,109 this month's harvest returned) carry no abstract and are not research: news, correspondence, editor highlights and obituaries. PubMed types them Journal Article and nothing else, so the publication-type filter cannot see them — across the whole index they are 22.7% of it and 35% of general, and 3,801 of them are JAMA. Every count below is over the full harvest and is inflated by this; the figure above is not.

General	Cancer	Immune	Bioinfo	Sysbio
483	192	88	314	32

Volume swing against the median month: sysbio +33%

Check the source before reading anything into this. Two benign causes account for most of these. bioinfo every January carries the Nucleic Acids Research database issue and runs high. And several journals simply deposit nothing in some months: across the 44 months read, 4 of 33 journals are absent from at least one, worst being Nat Mach Intell (20 months missing, 81 papers in total) and Bioinformatics (4 months missing, 2,905 papers in total). This month is missing Nat Mach Intell. Queried directly, PubMed returns zero for those journal-months too — the holes are in the source, reproduced exactly by re-running the harvest. So a swing here is not evidence that a field went quiet, and is not usually evidence that the pipeline broke either.

Largest contributors: Nucleic Acids Res 125, JAMA 112, Nature 106, Science 83, Bioinformatics 60, Lancet 50.

2. Net change on the thread curve

Thread	This month	Cumulative	Status
agentic-ai	+2	+11 / −1	forming
cell-therapy	+9 / −2	+185 / −54	forming
ctdna-mrd	+2	+63 / −22	forming
foundation-models	+4 / −6	+97 / −70	splitting
immunotherapy-resistance	+13 / −4	+201 / −73	forming
interception	+11 / −1	+116 / −22	forming
method-credibility	+2 / −13	+16 / −398	crisis
neoantigen-vaccines	+6	+75 / −30	forming
neuro-immune-microbiome-axis	+7 / −1	+152 / −15	forming
next-gen-checkpoints	+3	+64 / −35	splitting
single-cell-3d-genome	+2	+63 / −17	forming
suppressive-population-inversion	+5 / −1	+47 / −70	contested

Status is derived by rule, not asserted. Fewer than 5 cards → too-early; then refutes ≥ 2×supports → crisis; refutes > supports → contested; supports ≥ 2×refutes → forming; else splitting.

3. The evidence that moved it

The Virtual Cell Challenge is launched (40578317, Cell, 26 June) — framed in its own title as a Turing test for the virtual cell. Every dating claim this corpus makes about virtual-cell vocabulary should be measured from here, and note that the phrase was already in a Nature Biotechnology title two months earlier.

In the same month, the refutation of the thing being launched. 40407144 (Bioinformatics, 2 June): simple controls exceed the best deep learning algorithms for predicting genetic perturbations. Perturbation prediction is the challenge's central task. The corpus later treats "simple baselines match deep models" as a 2026 phenomenon and reads the July 2026 scaling nulls as the moment the premise broke; the claim was in print, in a title, three weeks before the challenge that assumed otherwise. 40434923 says the same thing for variant effects in the same month: conservation and solvent accessibility are almost all you need.

An autonomous agent is validated for clinical decisions. 40481323 (Nat Cancer): development and validation of an autonomous AI agent for clinical decision-making in oncology. This paragraph used to date the arrival by counting months forward from a February origin; data/threads.yml has voided that origin, and an agent is already the subject of a Nature Cancer piece in 2024-12 in any case. No interval replaces it. What the card is worth is what it contains — an agent with a validation study behind it, in an oncology journal — not its position in a chain.

The checkpoint roster turns over completely. June nominates Siglec-G/10, CMTM4, STS2, Piezo1, TTLL12, plus PSGL-1 as a phagocytosis checkpoint, TIM-3 and ENT1 — eight targets, all preclinical, none with an agent. Of May's five, one reappears. The templated block is at 31, near its peak. A field can be simultaneously at maximum output and producing nothing that will be tested.

4. Refutations

(a) Landing on a tracked thread

40457060: PD-1 receptor deficiency enhances CD30+ Treg function in melanoma — removing the checkpoint strengthens the suppressive population, which is the mechanism by which anti-PD-1 could make a subset of patients worse.

40482640 (Cell): microbiome mismatches from microbiota transplants produce persistent off-target metabolic and immunomodulatory effects. Filed against the neuro-immune-microbiome thread, and directly relevant to the faecal-transplant trials the thread later counts as supports.

Corrected 2026-08-27

Three more cards were listed here as refutations. None of them is one, and two of them were misread the same way — from the cell type named in the title rather than from the outcome reported in the abstract.

40202509, "MYO1F in neutrophils is required for the response to immune checkpoint blockade", was taken as a suppressive population being necessary for immunotherapy. The paper says tumour-associated neutrophils are a significant barrier, that MYO1F is what restrains their suppression, and that tumour-derived TGF-beta1 silencing it is what drives them into a STAT3-dependent immunosuppressive state that exhausts CTLs.

40146036 was taken as an inversion because a dendritic cell — presenting and stimulatory by convention — is the suppressor. Its outcome is the thread's claim verbatim: "moDC depletion limits Treg generation and improves the therapeutic outcome of RT". A newly identified suppressive population whose depletion helps supports the claim; only the identity of the population is new.

40335697 (Nature, oncogene aberrations drive medulloblastoma progression, not initiation) was read from its title as bad news for interception. What it does is re-time the driver events: the initiating lesions in group 3/4 medulloblastoma are large-scale chromosomal aberrations dated to the first gestational trimester, with single-gene oncogene hits arriving late and mostly subclonal. Its own conclusion is that single-cell technologies "can be applied for early detection and diagnosis of this fatal disease". It neither treats a precursor state nor shows interception failing, so it now sits on neither side of the ratio.

(b) Method credibility

40163891 (Blood): hallmarks of T-cell exhaustion and antigen experience are absent in multiple myeloma from diagnosis through maintenance. A negative result against a framework, published in the disease where the framework is most heavily invoked.

40578363 (Immunity): diet-derived LPS determines intestinal IgA induction and repertoire independently of the microbiota. It reassigns to diet what a large literature credits to microbes, which makes it both a method-credibility card and a warning about the microbiome arm of the neuro-immune thread.

40240607 (Nature): low overlap between transcription factor DNA binding and regulatory targets.

40195448 (Nat Med): sociodemographic biases in medical decision making by large language models — title-stated empirical evidence of failure, in a month whose clinical-AI counterweight is otherwise governance commentary.

5. Open items

Nothing carried forward.

6. Meetings

1 meeting with a written vault this month. These are read and summarised rather than harvested as abstracts, so they do not appear in the counts below.

• Galaxy and Bioconductor Community Conference 2025

No abstract book was deposited this month either — the vault above is built from transcripts and programmes rather than from a published abstract set. The abstract layer runs 2023-02 to 2026-08 and is seasonal by nature — meetings are annual, so most months are empty here and that is the expected shape, not a gap in the harvest.

7. Regulatory

Event	Count
trial-first-posted	516 (of which 463 genuinely oncology)
ema-authorisation	18
fda-approval	6

6 FDA actions — dated from each notice's own prose, not the table's posting date, which disagrees on 7 of 190 rows and once by enough to move an approval into the wrong month:

• 2025-06-03 — FDA approves darolutamide for metastatic castration-sensitive prostate cancer
• 2025-06-11 — FDA approves taletrectinib for ROS1-positive non-small cell lung cancer
• 2025-06-12 — FDA approves neoadjuvant and adjuvant pembrolizumab for resectable locally advanced head and neck squamous cell carcinoma
• 2025-06-12 — FDA approves mitomycin intravesical solution for recurrent low-grade intermediate-risk non-muscle invasive bladder cancer
• 2025-06-18 — FDA approves tafasitamab-cxix for relapsed or refractory follicular lymphoma
• 2025-06-23 — FDA grants accelerated approval to datopotamab deruxtecan-dlnk for EGFR-mutated non-small cell lung cancer

6 EMA oncology authorisations:

• 2025-06-23 — Yaxwer
• 2025-06-25 — Vevzuo
• 2025-06-26 — Denbrayce
• 2025-06-26 — Enwylma
• 2025-06-27 — Ziihera
• 2025-06-30 — Dazublys

8. Trade press

48 on-topic of 337 items across 3 feeds (n-stat 326, n-aacr-blog 7, n-acir 4). Headlines are selected by subject keyword, not recency — most of what these feeds carry is health policy and industry finance the corpus does not read.

• 2025-06-30 — Bloodlines Series: Mapping the Many Paths of Acute Myeloid Leukemia — n-aacr-blog
• 2025-06-30 — STAT+: HHS to continue funding cancer prevention, tracking work in states after uncertainty — n-stat
• 2025-06-30 — STAT+: AbbVie snaps up CAR-T company in a deal worth $2.1 billion — n-stat
• 2025-06-27 — Editors’ Picks, June 2025: Intercepting Pancreatic Cancer, Predicting Breast Cancer, and More — n-aacr-blog
• 2025-06-26 — Cancer Today’s Summer 2025 Issue: Expanded Use of Neoadjuvant Immunotherapy, PSMA-targeted Therapy in Prostate Cancer, and More — n-aacr-blog
• 2025-06-26 — STAT+: Generic cancer drugs are failing quality tests — n-stat
• 2025-06-26 — STAT+: States anxiously wait to find out if cancer tracking and prevention funding from CDC will be renewed — n-stat
• 2025-06-25 — A Bacterial Gut Punch to Enhance CAR T-cell Therapy — n-aacr-blog
• 2025-06-25 — Generic cancer drugs used around the world fail quality tests, investigation shows — n-stat
• 2025-06-25 — STAT+: DeepMind launches AlphaGenome, aiming to predict gene regulation from DNA sequence — n-stat
• 2025-06-25 — PSGL-1: a novel immune checkpoint for phagocytosis — n-acir
• 2025-06-24 — STAT+: Nuvalent says its ROS1-targeted drug shrank lung tumors in patients who failed other options — n-stat
• … and 36 more on-topic items.