---
tags:
- dna
- human_genome
---

# GENA-LM (gena-lm-bigbird-base-t2t)

GENA-LM is a Family of Open-Source Foundational Models for Long DNA Sequences.

GENA-LM models are transformer masked language models trained on human DNA sequence. 

`gena-lm-bigbird-base-t2t` follows the BigBird architecture and its HuggingFace implementation.

Differences between GENA-LM (`gena-lm-bigbird-base-t2t`) and DNABERT:
- BPE tokenization instead of k-mers;
- input sequence size is about 36000 nucleotides (4096 BPE tokens) compared to 512 nucleotides of DNABERT;
- pre-training on T2T vs. GRCh38.p13 human genome assembly.

Source code and data: https://github.com/AIRI-Institute/GENA_LM

Paper: https://academic.oup.com/nar/article/53/2/gkae1310/7954523

This repository also contains models that are finetuned on downstream tasks and models that are used in our [GENA-Web](https://dnalm.airi.net) web tool for genomic sequence annotation:
- splice sites prediction (branch [gena_web_spliceai](https://huggingface.co/AIRI-Institute/gena-lm-bigbird-base-t2t/tree/gena_web_spliceai))

## Examples

### Load pre-trained model
```python
from transformers import AutoTokenizer, BigBirdForMaskedLM

tokenizer = AutoTokenizer.from_pretrained('AIRI-Institute/gena-lm-bigbird-base-t2t')
model = BigBirdForMaskedLM.from_pretrained('AIRI-Institute/gena-lm-bigbird-base-t2t')
```


### How to load the model to fine-tune it on classification task
```python
from transformers import AutoTokenizer, BigBirdForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained('AIRI-Institute/gena-lm-bigbird-base-t2t')
model = BigBirdForSequenceClassification.from_pretrained('AIRI-Institute/gena-lm-bigbird-base-t2t')
```

## Model description
GENA-LM (`gena-lm-bigbird-base-t2t`) model is trained in a masked language model (MLM) fashion, following the methods proposed in the BigBird paper by masking 15% of tokens. Model config for `gena-lm-bigbird-base-t2t` is similar to the `google/bigbird-roberta-base`:

- 4096 Maximum sequence length
- 12 Layers, 12 Attention heads
- 768 Hidden size
- sparse config:
    - block size: 64
    - random blocks: 3
    - global blocks: 2
    - sliding window blocks: 3
- 32k Vocabulary size, tokenizer trained on DNA data.

We pre-trained `gena-lm-bigbird-base-t2t` using the latest T2T human genome assembly (https://www.ncbi.nlm.nih.gov/assembly/GCA_009914755.3/). The data was augmented by sampling mutations from 1000-genome SNPs (gnomAD dataset). Pre-training was performed for 1,070,000 iterations with batch size 256.

## Evaluation
For evaluation results, see our paper: https://academic.oup.com/nar/article/53/2/gkae1310/7954523

## Citation
```bibtex
@article{GENA_LM,
    author = {Fishman, Veniamin and Kuratov, Yuri and Shmelev, Aleksei and Petrov, Maxim and Penzar, Dmitry and Shepelin, Denis and Chekanov, Nikolay and Kardymon, Olga and Burtsev, Mikhail},
    title = {GENA-LM: a family of open-source foundational DNA language models for long sequences},
    journal = {Nucleic Acids Research},
    volume = {53},
    number = {2},
    pages = {gkae1310},
    year = {2025},
    month = {01},
    issn = {0305-1048},
    doi = {10.1093/nar/gkae1310},
    url = {https://doi.org/10.1093/nar/gkae1310},
    eprint = {https://academic.oup.com/nar/article-pdf/53/2/gkae1310/61443229/gkae1310.pdf},
}
```