ibm-research/biomed.rna.bert.110m.wced.multitask.v1
Biomedical foundational models for omics data. This package supports the development of foundation models for scRNA or for DNA data.
biomed-multi-omic enables development and testing of foundation models for DNA sequences and for RNA expression,
with modular model and training methods for pretraining and fine-tuning, controllable via a declarative no-code interface.
biomed-multi-omic leverages anndata, HuggingFace Transformers, PyTorchLighting and Hydra.
- 🧬 A single package for DNA and RNA Foundation models. scRNA pretraining on h5ad files or TileDB (eg CellXGene), DNA pretraining on reference human genome (GRCh38/hg38) and also variant imputed genome based on common SNPs available from GWAT catalog and ClinVar datasets.
- 🚀 Leverages latest open source tools: anndata, HuggingFace transformers and PyTorchLighting
- 📈 Zero-shot and finetuning support for diverse downstream tasks: (cell type annotation, perturbation prediction for scRNA, promoter prediction task and regulatory regions using Massively parallel reporter assays (MPRAs)
for DNA sequences)
- Novel pretraining strategies for scRNA and DNA implemented alongside existing methods to enable experimentation and comparison.
For details on how the models were trained, please refer to
the BMFM-RNA preprint.
Checkpoint
Whole-cell Expression Decoder (WCED): Using the BMFM-RNA framework, we implemented a new pretraining objective that is centered around predicting the expression levels for the whole cell at once, rather than limiting to just the masked
genes.
Multitask objectives: multi-label classification (cell type, tissue, tissue general), and an adversarial loss to unlearn donor ID.
WCED + Multitask: Trained first using WCED with random gene order and log-normalization, then fine-tuned with multitask objectives.
See section 2.3.3 of
the BMFM-RNA manuscript for more details.
Usage
Using
biomed.rna.bert.110m.wced.multitask.v1 requires the codebase
https://github.com/BiomedSciAI/biomed-multi-omicFor installation, please follow the
instructions on github.
RNA Inference
To get embeddings and predictions for scRNA data run:
1export MY_DATA_FILE=... # path to h5ad file with raw counts and gene symbols
2bmfm-targets-run -cn predict input_file=$MY_DATA_FILE working_dir=/tmp checkpoint=ibm-research/biomed.rna.bert.110m.wced.multitask.v1
For more details see the
RNA tutorials on github.
Citation
1@misc{dandala2025bmfmrnaopenframeworkbuilding,
2 title={BMFM-RNA: An Open Framework for Building and Evaluating Transcriptomic Foundation Models},
3 author={Bharath Dandala and Michael M. Danziger and Ella Barkan and Tanwi Biswas and Viatcheslav Gurev and Jianying Hu and Matthew Madgwick and Akira Koseki and Tal Kozlovski and Michal Rosen-Zvi and Yishai Shimoni and Ching-Huei Tsou},
4 year={2025},
5 eprint={2506.14861},
6 archivePrefix={arXiv},
7 primaryClass={q-bio.GN},
8 url={https://arxiv.org/abs/2506.14861},
9}
