README.md

---
tags:
- dna
---

# GENA-LM Fly 🪰 (gena-lm-bert-base-fly)

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

`gena-lm-bert-base-fly` is trained on drosophila genome.

## Model description
GENA-LM (`gena-lm-bert-base-fly`) model is trained with a masked language model (MLM) objective, following data preprocessing methods pipeline in the BigBird paper and by masking 15% of tokens. Model config for `gena-lm-bert-base-fly` is similar to the bert-base:

- 512 Maximum sequence length
- 12 Layers, 12 Attention heads
- 768 Hidden size
- 32k Vocabulary size

We pre-trained `gena-lm-bert-base-fly` on data obtained from Progressive Cactus alignment of 298 drosophilid species generated by [Kim et al.](https://www.biorxiv.org/content/10.1101/2023.10.02.560517v1), dataset source: [link](https://doi.org/10.5061/dryad.x0k6djhrd).
Pre-training was performed for 1,925,000 iterations with batch size 256 and sequence length was equal to 512 tokens. We modified Transformer to use [Pre-Layer normalization](https://arxiv.org/abs/2002.04745). We upload the checkpoint with the best loss on validation set.

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

Paper: https://www.biorxiv.org/content/10.1101/2023.06.12.544594

## Examples

### How to load pre-trained model for Masked Language Modeling
```python
from transformers import AutoTokenizer, AutoModel

tokenizer = AutoTokenizer.from_pretrained('AIRI-Institute/gena-lm-bert-base-fly')
model = AutoModel.from_pretrained('AIRI-Institute/gena-lm-bert-base-fly', trust_remote_code=True)

```

### How to load pre-trained model to fine-tune it on classification task
Get model class from GENA-LM repository:
```bash
git clone https://github.com/AIRI-Institute/GENA_LM.git
```

```python
from GENA_LM.src.gena_lm.modeling_bert import BertForSequenceClassification
from transformers import AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained('AIRI-Institute/gena-lm-bert-base-fly')
model = BertForSequenceClassification.from_pretrained('AIRI-Institute/gena-lm-bert-base-fly')
```
or you can just download [modeling_bert.py](https://github.com/AIRI-Institute/GENA_LM/tree/main/src/gena_lm) and put it close to your code.

OR you can get model class from HuggingFace AutoModel:
```python
from transformers import AutoTokenizer, AutoModel
model = AutoModel.from_pretrained('AIRI-Institute/gena-lm-bert-base-fly', trust_remote_code=True)
gena_module_name = model.__class__.__module__
print(gena_module_name)
import importlib
# available class names:
# - BertModel, BertForPreTraining, BertForMaskedLM, BertForNextSentencePrediction,
# - BertForSequenceClassification, BertForMultipleChoice, BertForTokenClassification,
# - BertForQuestionAnswering
# check https://huggingface.co/docs/transformers/model_doc/bert
cls = getattr(importlib.import_module(gena_module_name), 'BertForSequenceClassification')
print(cls)
model = cls.from_pretrained('AIRI-Institute/gena-lm-bert-base-fly', num_labels=2)
```

## Evaluation
For evaluation results, see our paper: https://www.biorxiv.org/content/10.1101/2023.06.12.544594


## Citation
```bibtex
@article{GENA_LM,
	author = {Veniamin Fishman and Yuri Kuratov and Maxim Petrov and Aleksei Shmelev and Denis Shepelin and Nikolay Chekanov and Olga Kardymon and Mikhail Burtsev},
	title = {GENA-LM: A Family of Open-Source Foundational DNA Language Models for Long Sequences},
	elocation-id = {2023.06.12.544594},
	year = {2023},
	doi = {10.1101/2023.06.12.544594},
	publisher = {Cold Spring Harbor Laboratory},
	URL = {https://www.biorxiv.org/content/early/2023/11/01/2023.06.12.544594},
	eprint = {https://www.biorxiv.org/content/early/2023/11/01/2023.06.12.544594.full.pdf},
	journal = {bioRxiv}
}

```