Biology subjects

Cheung, A. M.

Publications and source records attributed to Cheung, A. M..

3 recordsLinked to original sources

bioRxiv2026

scACORN: Context-engineered agent orchestration of specialized small language models for single-cell transcriptomic interpretation

Single-cell atlases now exceed 66 million cells, but turning a ranked expression profile and a free-form biological question into a reliable, evidence-grounded answer remains unsolved. Scaling a single model does not resolve this, because single-cell interpretation is a heterogeneous family of tasks whose correct answer depends on tissue, cohort, perturbation and annotation resolution. Here we present scACORN, an agentic alternative to monolithic single-cell language models that combines specialized small language models with context-engineered agent orchestration for their selection and composition at inference time. Each expert is built in two stages: domain-aligned contrastive adaptation fits a pretrained cell-to-text backbone to the transcriptomic geometry of a target dataset, and geometry-preserving specialization learns question-conditioned biological completions without eroding that geometry. A fixed orchestrating language model agent then selects and combines experts under a natural-language playbook that is itself optimized from textual feedback, with no gradient updates to the orchestrator. Across 10 Tabula Sapiens tissues, domain alignment raised transfer macro-F1 from 0.36 to 0.64 and Recall@5 from 0.87 to 0.97; specialized experts reached 0.89 mean exact-match annotation accuracy; and playbook optimization reduced unsupported gene citations from 14.5% to 3.5%. Our findings support specialization and orchestration as complementary responses to the heterogeneity and evidentiary demands of single-cell analysis.

bioinformatics↗

bioRxiv2022

General stain deconvolution of histopathology images with physics-guided deep learning

Advances have been made in the use of deep learning to extract quantitative and predictive information from digital pathology slides, yet many barriers remain before clinical translation and deployment. In particular, models need to be generalizable despite the wide variations in image characteristics due to inter-scanner variability and differences in slide preparation protocols. This has led to an interest in stain deconvolution methods that could correct for the variability in image appearances. However, most existing stain deconvolution methods were developed and validated on specific datasets and perform poorly on unseen data. We developed Physics-Guided Deep Image Prior network for Stain deconvolution (PGDIPS), a method that combines a novel optical physics model and a self-supervised deep neural network to perform deconvolution for various classes and any number of stains, without the need of training data. PGDIPS outperformed state-of-the-art approaches for the deconvolution of conventional stain combinations, enabled analysis of previously unsupported special stains, and provided superior interpretability by explicitly encoding representations for stain properties and the light transmittance/absorbance process. PGDIPS is publicly available as an end-to-end off-the-shelf tool that does not require data curation, domain knowledge or high computation power.

bioinformatics↗

bioRxiv2020

Structure of the complete, membrane-assembled COPII coat reveals a complex interaction network.

The COPII coat mediates Endoplasmic Reticulum (ER) to Golgi trafficking for thousands of proteins. Five essential coat proteins assemble at the ER into a characteristic two-layer architecture, which recruits cargo proteins whilst sculpting membrane carriers with diverse morphologies. How coat architecture drives membrane curvature whilst ensuring morphological plasticity is largely unknown, yet is central to understanding mechanisms of carrier formation. Here, we use an established reconstitution system to visualise the complete, membrane-assembled COPII coat with unprecedented detail by cryo-electron tomography and subtomogram averaging. We discover a network of interactions within and between coat layers, including multiple interfaces that were previously unknown. We reveal the physiological importance of these interactions using genetic and biochemical approaches. A newly resolved Sec31 C-terminal domain provides order to the coat and is essential to drive membrane curvature in cells. Moreover, a novel outer coat assembly mode provides a basis for coat adaptability to varying membrane curvatures. Furthermore, a newly resolved region of Sec23, which we term the L-loop, imparts coat stability and in part dictates membrane shape. Our results suggest these interactions collectively contribute to coat organisation and membrane curvature, providing a structural framework to understand regulatory mechanisms of COPII trafficking and secretion.

cell biology↗