National University of Singapore

Protein perturbation model conditioning substitution-effect prediction on a learned protein-level regime coordinate over 202 million sequences.

Sequence-conditioned generative framework that reconstructs missing prostate MRI contrasts and restores artefact-degraded acquisitions.

Infers gene-centered chromatin interactions from bulk RNA-seq alone, mapping 3D genome changes across 12,347 tumor and normal transcriptomes.

EEG foundation model that corrects low-frequency bias by reconstructing band-standardized time-frequency targets. State of the art on 24 of 41 tasks.

Atomic-level refinement of RNA 3D structures, using geometric attention networks to guide physics-based Monte Carlo sampling and L-BFGS optimization.

Histopathology model reconstructing tissue-wide single-cell gene expression from H&E slides, using sparse TMA measurements as molecular anchors.

370M-parameter ligand-conditioned discrete diffusion model that co-designs protein sequence and structure under explicit small-molecule constraints.

Paired-sequence protein language model that jointly encodes two interacting chains to predict interactions, binding affinity, and interface contacts.

Enzyme backbone design model that adds substrate and catalytic-site control to a pretrained SE(3) flow-matching generator via a lightweight adapter.

Single-cell foundation model domain-adapting Llama-3.1-8B on 1.3M gastric cancer cells with gene-family cell sentences instead of ranked-gene order.

Diploid de novo genome assembly that scores assembly-graph edges with a graph neural network and reconstructs phased haplotypes by beam search.

Protein multi-conformation predictor that scores per-residue flexibility, then masks MSA columns to steer AlphaFold 2 toward alternative states.

Self-supervised medical imaging foundation model pretrained on 3.3 million CT, X-ray, ultrasound, pathology, OCT, fundus, and dermoscopy images.

EEG foundation model pairing a decoupled time-frequency tokenizer with a multi-scale state-space encoder for generalization under distribution shift.

Sequence-only predictor of protein stability change on point mutation, scoring both ddG and melting temperature shift without any input structure.

Small-molecule foundation model pairing a graph encoder with a Transformer SMILES decoder so compounds can be optimized directly in encoding space.

Single-cell foundation model that fuses scRNA-seq profiles with text, pairing a cell encoder with an LLM for cell annotation and clustering.

Ab initio RNA 3D structure prediction from a single sequence, using a composite-likelihood language model and a denoising end-to-end structure module.

Multimodal LLM unifying 12-lead ECG time series, ECG images, and text for grounded, clinician-aligned electrocardiogram interpretation.

3D molecule generation that writes a valid 1D SELFIES string with a pretrained language model, then predicts its conformer with a diffusion module.

Medical vision-language model that unifies image comprehension and generation in one autoregressive transformer via heterogeneous LoRA adapters.

Nanopore basecaller extending Bonito to a six-letter alphabet, reading the unnatural bases Ds and Px alongside canonical A, T, C and G.

Cell-type-specific gene expression prediction from DNA sequence, mapping Enformer epigenomic features to pseudobulk expression for cell-resolved TWAS.

De novo peptide sequencing model that retrieves a similar peptide-spectrum match from a database and fuses it into transformer decoding.

Brain-dynamics foundation model for resting-state fMRI, adapting the Joint-Embedding Predictive Architecture with brain gradient positioning.

SAM2-based foundation model that segments 2D and 3D medical images by treating volumes and image sets as video object tracking.

Ophthalmic imaging foundation model pretrained on 2.78M images across 11 modalities for diagnosis, prognosis, and visual question answering.

Lightweight mixture-of-experts medical vision-language model routing visual question answering and image classification to domain-specific experts.

Tokenized graph transformer that embeds longitudinal brain functional connectomes from fMRI for interpretable neurodegenerative disease diagnosis.

Self-supervised vision foundation model pretrained on 1.3M medical images via second-order graph matching, for segmentation and classification.

Chromatin interaction prediction from DNA sequence alone, calling CTCF-, RNA Pol II- and Hi-C-associated loops between open chromatin regions.