Research Radar — 2026-08-08
Methods & AI
Computational
Spatial multi-omics and single-cell transcriptomics uncover senescence-associated cellular programs during colon adenoma to cancer progression
bioRxiv (cancer_biology) Published 2026-08-07 Preprint DOI: 10.64898/2026.08.06.743310v1
spatial multi-omics single-cell transcriptomics cellular senescence colon cancer
Summary: The study uses spatial multi-omics and single-cell transcriptomics to examine senescence-associated cellular programs across progression from colon adenoma to cancer.
Why it matters: Linking senescence programs to spatial and single-cell changes may clarify how cellular states evolve during colorectal tumorigenesis.
Why for Yiru: It directly combines spatial multi-omics with single-cell analysis in a cancer-progression setting relevant to tumor-microenvironment research.
MaskTalk: cell-identity-gated spatial lag for target-aware cell-cell communication inference in high-resolution spatial transcriptomics
bioRxiv (bioinformatics) Published 2026-08-06 Preprint DOI: 10.64898/2026.08.02.741948v1
spatial transcriptomics cell-cell communication spatial lag computational biology
Summary: MaskTalk presents a cell-identity-gated spatial-lag approach for target-aware inference of cell-cell communication from high-resolution spatial transcriptomics.
Why it matters: Target-aware communication inference can help distinguish spatially supported signaling relationships from broad cell-proximity associations.
Why for Yiru: The method is directly relevant to computational analysis of spatially organized tumor-microenvironment interactions.
Learning with Recurrence Geometric AI in Spatial Transcriptomics
bioRxiv (bioinformatics) Published 2026-08-06 Preprint DOI: 10.64898/2026.08.02.742287v1
spatial transcriptomics geometric AI machine learning
Summary: This work applies a recurrence-based geometric AI framework to learning from spatial transcriptomics data.
Why it matters: Geometric learning may provide a way to model spatial structure while retaining relationships among transcriptomic measurements.
Why for Yiru: It fits the intersection of spatial transcriptomics and machine-learning methods central to computational TME analysis.
Evaluating Lightweight and Full Fine-Tuning Strategies Against Classical Machine Learning for Protein Function Prediction
bioRxiv (bioinformatics) Published 2026-08-07 Preprint DOI: 10.64898/2026.08.02.737389v1
protein function prediction fine-tuning machine learning protein models
Summary: The study compares lightweight and full fine-tuning strategies with classical machine-learning methods for protein function prediction.
Why it matters: Comparing adaptation strategies against classical baselines helps calibrate the practical value and computational cost of protein-prediction models.
Why for Yiru: The benchmark is relevant to selecting efficient machine-learning strategies for biological sequence and protein-function problems.
Biomedical discoveries
Biomedicine
APOE+ macrophages and POSTN+CAFs form immune-exclusion niche driving immunotherapy resistance in clear-cell renal carcinoma
Science Advances Published 2026-08-07 Research Article DOI: 10.1126/sciadv.aed0733
clear-cell renal carcinoma macrophages cancer-associated fibroblasts immune exclusion immunotherapy resistance
Summary: The article describes an immune-exclusion niche formed by APOE-positive macrophages and POSTN-positive cancer-associated fibroblasts in clear-cell renal carcinoma.
Why it matters: The macrophage–CAF niche provides a cellular framework for studying microenvironmental resistance to immunotherapy.
Why for Yiru: It directly connects myeloid and stromal states to spatial tumor-microenvironment analysis and immunotherapy response.
Spatial multi-omics reveals targetable immunosuppressive macrophage T-cell interactions in human AML bone marrow
bioRxiv (cancer_biology) Published 2026-08-07 Preprint DOI: 10.64898/2026.08.07.743431v1
acute myeloid leukemia spatial multi-omics macrophages T cells bone marrow
Summary: The study uses spatial multi-omics to investigate immunosuppressive macrophage–T-cell interactions in human acute myeloid leukemia bone marrow.
Why it matters: Spatially resolving macrophage–T-cell interactions can identify microenvironmental relationships with therapeutic relevance in AML.
Why for Yiru: This is a direct example of spatial multi-omics applied to immune-cell interactions in a clinically important tumor niche.
Immune-metabolic PET/MRI uncovers microenvironmental reprogramming under combined immunotherapy and anti-angiogenic therapy
bioRxiv (cancer_biology) Published 2026-08-07 Preprint DOI: 10.64898/2026.08.06.743278v1
PET/MRI tumor microenvironment immunotherapy anti-angiogenic therapy metabolism
Summary: The work uses immune-metabolic PET/MRI to examine microenvironmental reprogramming during combined immunotherapy and anti-angiogenic therapy.
Why it matters: A non-invasive view of treatment-associated microenvironmental change could help connect therapy response with immune and metabolic states.
Why for Yiru: The imaging approach complements spatial and molecular TME measurements for studying treatment response.
T cell–nanodrug conjugates synchronize vascular normalization and immune activation for solid tumor therapy
Science Advances Published 2026-08-07 Research Article DOI: 10.1126/sciadv.adp2955
T cells nanodrug conjugates vascular normalization immune activation solid tumors
Summary: The article investigates T cell–nanodrug conjugates designed to synchronize vascular normalization with immune activation in solid-tumor therapy.
Why it matters: Coordinating vascular and immune effects addresses two linked constraints on treatment delivery and activity in solid tumors.
Why for Yiru: It connects immune-cell engineering with tumor-microenvironment remodeling and translational immunotherapy.
CAFs shape the immunosuppressive microenvironment of pancreatic cancer through the Lin28b-STING Axis
Nature Published 2026-08-07 Research Article DOI: 10.1038/s41467-026-76495-3
pancreatic cancer cancer-associated fibroblasts immunosuppression Lin28b STING
Summary: The article examines how cancer-associated fibroblasts shape the immunosuppressive pancreatic-cancer microenvironment through the Lin28b–STING axis.
Why it matters: A defined fibroblast signaling axis offers a mechanistic entry point for studying stromal control of pancreatic-cancer immunity.
Why for Yiru: It provides a focused stromal-signaling example for spatial and single-cell models of the pancreatic tumor microenvironment.
alpha2,3-sialylation on human naive T cells restrains bispecific engager-mediated anti-tumor immunity
bioRxiv (immunology) Published 2026-08-06 Preprint DOI: 10.64898/2026.08.02.742298v1
T cells sialylation bispecific engagers anti-tumor immunity
Summary: The study reports that alpha2,3-sialylation on human naive T cells restrains bispecific engager-mediated anti-tumor immunity.
Why it matters: The finding identifies a post-translational cell-surface feature that may influence the activity of bispecific immune therapies.
Why for Yiru: It links T-cell state and glycosylation to therapeutic immune activation, relevant to translational immunotherapy research.
Cross-disciplinary watchlist
Other Fields
Direct visualization of RNA virus and evaluation of drug effect in living cells by catalytically transformable DNA cube
Science Advances Published 2026-08-07 Research Article DOI: 10.1126/sciadv.aee2515
RNA viruses live-cell imaging DNA nanotechnology drug evaluation
Summary: A catalytically transformable DNA cube is used to visualize RNA virus and evaluate drug effects in living cells.
Why it matters: Live-cell visualization can provide a direct readout for studying viral infection and treatment response.
Why for Yiru: The programmable nucleic-acid platform illustrates a route to quantitative biological measurements for computationally informed experiments.
Stability of c-Myc protein differentiates Ras oncogene addiction and MAPK pathway dependency in Ras-mutant multiple myeloma
bioRxiv (cancer_biology) Published 2026-08-07 Preprint DOI: 10.64898/2026.08.06.743109v1
c-Myc Ras MAPK pathway multiple myeloma oncogene addiction
Summary: The study examines how c-Myc protein stability differentiates Ras oncogene addiction from MAPK pathway dependency in Ras-mutant multiple myeloma.
Why it matters: Separating oncogene addiction from pathway dependency may help explain heterogeneous responses to targeted perturbations.
Why for Yiru: It offers a mechanistic cancer-biology reference for interpreting signaling dependencies and treatment response.
KlinkPPI: Single Point of Access to Protein-Protein Interactions Across Databases
bioRxiv (bioinformatics) Published 2026-08-06 Preprint DOI: 10.64898/2026.08.02.742057v1
protein-protein interactions databases bioinformatics data integration
Summary: KlinkPPI provides a single point of access to protein-protein interaction information across databases.
Why it matters: Centralized access can reduce friction when gathering interaction evidence for biological analysis.
Why for Yiru: Protein-interaction resources are useful for connecting molecular mechanisms with network and systems-biology models.
Ribosomal protein bL27 protects translating ribosomes from tmRNA-SmpB
Science Advances Published 2026-08-07 Research Article DOI: 10.1126/sciadv.aef7696
ribosomes translation bL27 tmRNA-SmpB bacterial quality control
Summary: The study investigates how ribosomal protein bL27 protects translating ribosomes from tmRNA-SmpB.
Why it matters: The work addresses a core mechanism of ribosome protection and bacterial translation quality control.
Why for Yiru: It provides a mechanistic reference for interpreting translation-related phenotypes in systems biology and protein-function studies.