Biology subjects

Huang, J.-J.

Publications and source records attributed to Huang, J.-J..

3 recordsLinked to original sources

bioRxiv2026

Morpho-FM: spatial molecular reconstruction from routine H&E histology using transcriptomic foundation-model priors

Routine haematoxylin and eosin (H&E) histology captures tissue architecture at clinical scale, but lacks a direct molecular readout of the transcriptional programmes that organise tumour epithelium, stroma, vasculature and immune compartments. Spatial transcriptomics provides this context, yet cost, workflow complexity and sparse sampling limit routine use. Most existing histology-to-expression models are trained de novo on small paired cohorts and therefore remain weakly constrained when extrapolating from sparse measurements to dense, tissue-wide molecular maps. Here we introduce Morpho-FM, a weakly supervised framework that predicts spatial gene expression from routine H&E whole-slide images by conditioning a pretrained single-cell transcriptomic foundation-model prior on local histological neighbourhoods. A lightweight morphology-to-transcriptome adapter maps cached whole-slide histology features into a transcriptomic decoder, enabling prediction at measured locations, dense full-section reconstruction, and re-aggregation to the original measurement support. Across harmonized prostate cancer benchmarks, Morpho-FM achieved the strongest overall performance among five representative methods, reaching mean per-gene Pearson correlations of 0.286 in rotating single-slide evaluation and 0.298 in multi-slide held-out validation. The framework reproduced this advantage across kidney cancer sections, achieved a mean correlation of 0.210 across 56 directed single-slide evaluations and retained measurable predictive signal after external transfer to clear-cell renal cell carcinoma sections. Controlled ablation analyses identified pretrained transcriptomic initialization as a reproducible source of performance gain exceeding that attributable to changes in the histology feature backbone. Beyond predictive accuracy benchmarks, Morpho-FM recovered ERBB2-enriched tumour compartments, boundary-associated molecular gradients, and annotation-aligned tissue domains across Xenium and HER2ST breast cancer datasets. Together, these results support transcriptomic foundation-model priors as an effective constraint for morphology-conditioned molecular decoding and demonstrate the potential of Morpho-FM to extend spatial transcriptomic insight across routine pathology sections.

bioinformatics↗

bioRxiv2026

Progressive matrix stiffening of tyramine-modified silk fibroin hydrogels governs stage-specific pulmonary fibroblast activation

Fibrosis is a progressive and often fatal pathological process characterized by excessive extracellular matrix deposition, tissue stiffening, and irreversible organ dysfunction. Effective antifibrotic therapies remain limited by the lack of in vitro models that recapitulate the full spectrum of fibrotic disease progression. Here, we leverage tyramine-modified silk fibroin (SF-TA) hydrogels to investigate normal human lung fibroblasts (NHLF) responses to progressively stiffening environments relevant to pulmonary fibrosis. Two hydrogel formulations with distinct stiffening profiles over 14 days were prepared: a gradual-stiffening 0% SF-TA formulation reaching [~]20 kPa, and a rapidly stiffening 50% SF-TA formulation reaching [~]60 kPa. NHLFs were cultured on both formulations, with and without TGF{beta} (5 ng/mL), for 14 days and assessed for viability, metabolic activity, cytokine and collagen secretion, cytoskeletal organization, and mechanotransductive gene expression. The 0% SF-TA hydrogels drove sustained fibroblast proliferation and elevated secretion of IL-6, IL-8, and MCP-1, consistent with early inflammatory fibrosis. The 50% SF-TA hydrogels induced a metabolic plateau without senescence, suppressed inflammatory cytokine secretion, and, in the presence of TGF{beta}, led to significant upregulation of ACTA2 and CTGF, alongside -SMA stress fiber incorporation, consistent with established myofibroblast persistence. Both conditions produced comparable secreted collagen output by day 14. Together, these findings establish dynamically stiffening SF-TA hydrogels as a tunable platform for investigating stage-dependent fibroblast activation and mechanobiological progression in fibrosis.

bioengineering↗

bioRxiv2025

Neural coding for tactile motion: Scanning speed or temporal frequency?

Humans effortlessly perceive the speed of an object moving across their fingers, but how the brain encodes this information, especially across the hierarchical stages in the primary somatosensory cortex, remains unclear. This study thus investigated coding schemes, including rate and temporal codes, for tactile motion speed in macaque S1 areas 3b, 1, and 2. Extracellular electrophysiology recorded single-unit activities when a rotating sinusoidal grating ball of a fixed spatial period (wavelength of 1, 2, or 4 mm) was presented on the fingerpad at various speeds (20-320 mm/s). The results showed that the rate code was commonly employed to differentiate the stimulus scanning speed, spatial period, and scanning direction across S1 regions. In contrast, the temporal code was used to faithfully represent the stimulus temporal frequency, which was defined as the speed divided by the spatial period. Notably, area 3b had a wider range of frequency responses than did areas 1 and 2. These findings demonstrate that S1 uses both rate and temporal codes to encode distinct aspects of tactile motion. Future research should investigate how temporal patterns in S1 neuronal activity are potentially transformed and utilized in downstream somatosensory areas to form tactile motion perception and guide perceptual decisions.

neuroscience↗