Latest & Follow-up · English edition

Frontier One

New research, explained from first principles. These are full translations, with the original tables, equations, figures, sources and limitations retained.

16 complete English articles. Translation coverage · English RSS · 한국어 전체 NEWS

LTRI robotic synthesis, optical measurements and virtual formulation predictions — Original Figure 1 Zhao et al. · Nature Communications · Figure 1 · Source · CC BY-NC-ND 4.0

AI connects nanocrystal synthesis to photovoltaic process development

Room-temperature LTRI chemistry, robotic experiments, machine learning and 1 L flow synthesis meet device testing. Separate estimated costs from manufacturing costs and the three stability tests.

Read the full article ↗

MCCv allele-specific local architecture and distant regulatory connections — Original Figure 1 Hamley et al. · Nature Genetics · Figure 1 · Source · CC BY 4.0

MCCv traces a disease-associated DNA change to its regulatory effects

MCCv combines short-range chromatin contacts with allele-specific long-range mapping. Read the 405-target study, the SESN3 mechanism and the limits of causality, licensing and generalization.

Read the full article ↗

Pediatric AML population abundance and transcriptomic trajectories between diagnosis and relapse — Original Figure 2 NajafPanah et al. · Nature Communications · Figure 2 · Source · CC BY-NC-ND 4.0

Can rare leukemia-cell states at diagnosis improve relapse-risk prediction?

A pediatric AML study combines inferred resistant-cell populations with genetic risk and post-induction MRD. Separate prediction, survival endpoints and preclinical drug response from clinical benefit.

Read the full article ↗

An independent teaching diagram of I-O-T-H responsibilities — Independent teaching diagram · Not experimental data JJo · independent educational illustration · Source · CC BY 4.0

An LLM builds a complementary algorithm team instead of one universal answer

Read LACE’s I-O-T-H contract, complementary heuristic evolution and MILP selection. Separate the 0.945 benchmark score, reproducibility materials and operational claims.

Read the full article ↗

Original Figure 1: controls and metric calibration through DRF — Original Figure 1 Miller et al. · Nature Biotechnology · Figure 1 · Source · CC BY 4.0

Did the virtual-cell model fail, or did the benchmark miss the difference?

Calibration across 14 datasets and 18 metrics reframes perturbation prediction. Read positive controls, DRF, additive-baseline saturation and limits of generalization.

Read the full article ↗

Independent teaching sketch of background returns behind glass — Independent teaching diagram · Not experimental data JJo · independent educational illustration · Source · CC BY 4.0

Can depth priors stop a robot from seeing a glass door as empty space?

GlassRecon aligns a monocular structural prior with RGB-D measurements through local RANSAC. Read the 917-image evaluation, geometric ground truth and navigation limits.

Read the full article ↗

Original SynthID Bio Figure 1: sequence generation and detection above, and watermarking of predicted AlphaFold 3 coordinates below — Original Figure 1 · two different watermarking workflows David Stutz et al. · Nature (2026) · Figure 1 · Source · CC BY 4.0 · original paper

Can AI-designed proteins carry an invisible indication of their origin?

Read SynthID Bio through its original figures, binding experiments and worked examples. Separate sequence from digital structure watermarks, selected-design detection from general detection, and provenance from safety.

Read the full article ↗

Background photograph of repeated circuits on an ordinary silicon wafer; this is not the BCN wafer investigated in the paper — Background photograph · not the experimental BCN wafer Rob Bulmahn · 2020 · Wikimedia Commons · Source · CC BY 2.0

One-atom-thick BCN brings p-type transistors to wafer-scale material

Understand monolayer boron carbon nitride, holes, mobility and on/off ratio through a worked example and explanatory visuals. Separate wafer-scale material and transistor measurements from complete CMOS manufacturing.

Read the full article ↗

Original Ataraxos Figure 2: a Stratego board viewed from the red side with opposing blue identities concealed, followed by a battle example — Original Figure 2 · Stratego rules illustration Samuel Sokota et al. · Nature (2026) · Figure 2 · Source · CC BY 4.0 · original paper

How Ataraxos makes decisions when an opponent’s pieces are hidden

Read Ataraxos through the original Stratego board and game graphs. Understand hidden information, beliefs, self-play and the 15–1–4 result without confusing game performance with real-world decision-making.

Read the full article ↗

A 70-cell SOEC, alternating operating directions and voltage histories — Original Figure 1 Bilalis et al. · Nature Communications · Figure 1 · Source · CC BY-NC-ND 4.0

A 25,000-hour SOEC stack: durability through operating control

A 70-cell SOEC ran for 25,011 hours, with 0.05% per 1,000 hours reported in a later 22,268-hour window. Read the evidence, operating mechanism and limits.

Read the full article ↗

Round-specific intervention-arm analysis populations in NHS-Galleri — Original Figure 1 Neal et al. · Nature Medicine · Figure 1 · Source · CC BY 4.0

NHS-Galleri: high specificity is not proof of clinical benefit

Read NHS-Galleri’s secondary performance analysis: high specificity, conditional cancer-origin accuracy and the unmet randomized primary endpoint are different findings.

Read the full article ↗

The NPU-Sparrow airframe, tail-morphing linkage and wing-incidence mechanism — Original Figure 2 Cao et al. · Nature Communications · Figure 2 · Source · CC BY-NC-ND 4.0

NPU-Sparrow: a changing tail role expands robotic flight

NPU-Sparrow coordinates tail morphing and wing incidence. Understand its 4.5–13.2 m/s range, 375°/s peak roll rate, smaller turns and remaining validation gaps.

Read the full article ↗

Original Figure 1 comparing theoretical energy, sulfur compounds, and the Li₂S-to-S₂Cl₂ reaction concept — Original Figure 1 · theoretical comparison and reaction concept Nan Zhang et al. · Nature Energy · 2026 · Figure 1 · Source · Original copyright retained; limited reproduction for figure commentary

Can Sulfur Exchange One More Electron in a Battery?

A beginner’s guide to three-electron sulfur chemistry, voltage and capacity, electrode-level specific energy, and the limits of laboratory battery results.

Read the full article ↗

imec’s 2024 High-NA background image showing early CAR and MOR exposure patterns — Official background image · early High-NA patterning, 2024 ©imec · 2024 · Figure 1 · Source · imec press and presentation image terms

Why Smaller Chips Need Larger EUV Photomasks

ASML and TSMC’s large-mask roadmap explained through High-NA optics, exposure fields, stitching, and the separate 2030, 2031, and 2033 milestones.

Read the full article ↗

A blastocyst micrograph showing the internal cell cluster and cavity — Original institutional photograph · early human-embryo research Štěpán Jeřábek · Columbia University / IOCB Prague · Source · Institutional material; original copyright retained

Why Accurate Embryo Editing Does Not Establish Safety

Human embryo base editing explained: DNA, blastocysts, mosaicism, unintended changes, and why successful editing does not establish reproductive safety.

Read the full article ↗

Complete original Figure 2 and caption: approximately 28 μm cross-section and a 30 by 10 cm freestanding film — Original Figure 2 · film, cross-section, and elemental maps Chunya Lou et al. · Nature Communications · 2026 · Figure 2 · Source · CC BY-NC-ND 4.0 · noncommercial, no derivatives

Why Solid-State Batteries Need Thin Electrolyte Films

Understand sulfide electrolyte films, thickness and resistance, pouch-cell cycling, and the gap between useful battery materials and repeatable manufacturing.

Read the full article ↗