yang0104/TropicBERT

MIMIC is a multimodal encoder–decoder foundation model of the central dogma, trained jointly over DNA, RNA, and protein together with a range of structural and functional tracks. A single model embeds any subset of modalities into a shared representation space and generates any modality conditioned…

Foundation models for genomics and transcriptomics pretrained on 3,000+ human genomes and 850+ diverse species, enabling chromatin accessibility prediction, splice site detection, and promoter classification across multiple model scales (InstaDeep, NVIDIA & TUM, Nature Methods 2023)

This repository contains the model presented in D3LM: A Discrete DNA Diffusion Language Model for Bidirectional DNA Understanding and Generation.

CENO-1B-1m is a checkpoint of the CENO base DNA foundation model (1M context (stage 4)). It is a plain causal language model over genomic sequence on a Nemotron-H Mamba/Attention/MoE hybrid backbone, with no MSA inputs.