Instructions to use multimolecule/deltasplice-human with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MultiMolecule
How to use multimolecule/deltasplice-human with MultiMolecule:
pip install multimolecule
from multimolecule import AutoModel, AutoTokenizer tokenizer = AutoTokenizer.from_pretrained("multimolecule/deltasplice-human") model = AutoModel.from_pretrained("multimolecule/deltasplice-human") inputs = tokenizer("UAGCUUAUCAGACUGAUGUUGA", return_tensors="pt") outputs = model(**inputs) embeddings = outputs.last_hidden_state - Notebooks
- Google Colab
- Kaggle
File size: 17,145 Bytes
c16a7a1 7a97fb2 c16a7a1 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 | ---
datasets:
- multimolecule/deltasplice-human
library_name: multimolecule
license: agpl-3.0
pipeline: splice-site
pipeline_tag: other
tags:
- Biology
- RNA
- Splicing
- rna
widget:
- example_title: microRNA 21
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: UAGCUUAUCAGACUGAUGUUGA
- example_title: microRNA 146a
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: UGAGAACUGAAUUCCAUGGGUU
- example_title: microRNA 155
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: UUAAUGCUAAUCGUGAUAGGGGUU
- example_title: RNA component of mitochondrial RNA processing endoribonuclease
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: GGUUCGUGCUGAAGGCCUGUAUCCUAGGCUACACACUGAGGACUCUGUUCCUCCCCUUUCCGCCUAGGGGAAAGUCCCCGGACCUCGGGCAGAGAGUGCCACGUGCAUACGCACGUAGACAUUCCCCGCUUCCCACUCCAAAGUCCGCCAAGAAGCGUAUCCCGCUGAGCGGCGUGGCGCGGGGGCGUCAUCCGUCAGCUCCCUCUAGUUACGCAGGCAGUGCGUGUCCGCGCACCAACCACACGGGGCUCAUUCUCAGCGCGGCUGUAAAAAAAAA
- example_title: 7SK small nuclear RNA
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: GGAUGUGAGGGCGAUCUGGCUGCGACAUCUGUCACCCCAUUGAUCGCCAGGGUUGAUUCGGCUGAUCUGGCUGGCUAGGCGGGUGUCCCCUUCCUCCCUCACCGCUCCAUGUGCGUCCCUCCCGAAGCUGCGCGCUCGGUCGAAGAGGACGACCAUCCCCGAUAGAGGAGGACCGGUCUUCGGUCAAGGGUAUACGAGUAGCUGCGCUCCCCUGCUAGAACCUCCAAACAAGCUCUCAAGGUCCAUUUGUAGGAGAACGUAGGGUAGUCAAGCUUCCAAGACUCCAGACACAUCCAAAUGAGGCGCUGCAUGUGGCAGUCUGCCUUUCUUUU
- example_title: telomerase RNA component
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: GGGUUGCGGAGGGUGGGCCUGGGAGGGGUGGUGGCCAUUUUUUGUCUAACCCUAACUGAGAAGGGCGUAGGCGCCGUGCUUUUGCUCCCCGCGCGCUGUUUUUCUCGCUGACUUUCAGCGGGCGGAAAAGCCUCGGCCUGCCGCCUUCCACCGUUCAUUCUAGAGCAAACAAAAAAUGUCAGCUGCUGGCCCGUUCGCCCCUCCCGGGGACCUGCGGCGGGUCGCCUGCCCAGCCCCCGAACCCCGCCUGGAGGCCGCGGUCGGCCCGGGGCUUCUCCGGAGGCACCCACUGCCACCGCGAAGAGUUGGGCUCUGUCAGCCGCGGGUCUCUCGGGGGCGAGGGCGAGGUUCAGGCCUUUCAGGCCGCAGGAAGAGGAACGGAGCGAGUCCCCGCGCGCGGCGCGAUUCCCUGAGCUGUGGGACGUGCACCCAGGACUCGGCUCACACAUGC
- example_title: vault RNA 2-1
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: CGGGUCGGAGUUAGCUCAAGCGGUUACCUCCUCAUGCCGGACUUUCUAUCUGUCCAUCUCUGUGCUGGGGUUCGAGACCCGCGGGUGCUUACUGACCCUUUUAUGCAA
- example_title: brain cytoplasmic RNA 1
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: GGCCGGGCGCGGUGGCUCACGCCUGUAAUCCCAGCUCUCAGGGAGGCUAAGAGGCGGGAGGAUAGCUUGAGCCCAGGAGUUCGAGACCUGCCUGGGCAAUAUAGCGAGACCCCGUUCUCCAGAAAAAGGAAAAAAAAAAACAAAAGACAAAAAAAAAAUAAGCGUAACUUCCCUCAAAGCAACAACCCCCCCCCCCCUUU
- example_title: HIV-1 TAR-WT
pipeline_tag: splice-site
sequence_type: ncRNA
task: splice-site
text: GGUCUCUCUGGUUAGACCAGAUCUGAGCCUGGGAGCUCUCUGGCUAACUAGGGAACC
- example_title: prion protein (Kanno blood group)
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AUGGCGAACCUUGGCUGCUGGAUGCUGGUUCUCUUUGUGGCCACAUGGAGUGACCUGGGCCUCUGC
- example_title: interleukin 10
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AUGCACAGCUCAGCACUGCUCUGUUGCCUGGUCCUCCUGACUGGGGUGAGGGCC
- example_title: Zaire ebolavirus
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AAUGUUCAAACACUUUGUGAAGCUCUGUUAGCUGAUGGUCUUGCUAAAGCAUUUCCUAGCAAUAUGAUGGUAGUCACAGAGCGUGAGCAAAAAGAAAGCUUAUUGCAUCAAGCAUCAUGGCACCACACAAGUGAUGAUUUUGGUGAGCAUGCCACAGUUAGAGGGAGUAGCUUUGUAACUGAUUUAGAGAAAUACAAUCUUGCAUUUAGAUAUGAGUUUACAGCACCUUUUAUAGAAUAUUGUAACCGUUGCUAUGGUGUUAAGAAUGUUUUUAAUUGGAUGCAUUAUACAAUCCCACAGUGUUAU
- example_title: SARS coronavirus
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AUGUUUAUUUUCUUAUUAUUUCUUACUCUCACUAGUGGUAGUGACCUUGACCGGUGCACCACUUUUGAUGAUGUUCAAGCUCCUAAUUACACUCAACAUACUUCAUCUAUGAGGGGGGUUUACUAUCCUGAUGAAAUUUUUAGAUCAGACACUCUUUAUUUAACUCAGGAUUUAUUUCUUCCAUUUUAUUCUAAUGUUACAGGGUUUCAUACUAUUAAUCAUACGUUUGACAACCCUGUCAUACCUUUUAAGGAUGGUAUUUAUUUUGCUGCCACAGAGAAAUCAAAUGUUGUCCGUGGUUGGGUUUUUGGUUCUACCAUGAACAACAAGUCACAGUCGGUGAUUAUUAUUAACAAUUCUACUAAUGUUGUUAUACGAGCAUGUAACUUUGAAUUGUGUGACAACCCUUUCUUUGCUGUUUCUAAACCCAUGGGUACACAGACACAUACUAUGAUAUUCGAUAAUGCAUUUAAAUGCACUUUCGAGUACAUAUCU
- example_title: insulin
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AUGGCCCUGUGGAUGCGCCUCCUGCCCCUGCUGGCGCUGCUGGCCCUCUGGGGACCUGACCCAGCCGCAGCCUUUGUGAACCAACACCUGUGCGGCUCACACCUGGUGGAAGCUCUCUACCUAGUGUGCGGGGAACGAGGCUUCUUCUACACACCCAAGACCCGCCGGGAGGCAGAGGACCUGCAGGUGGGGCAGGUGGAGCUGGGCGGGGGCCCUGGUGCAGGCAGCCUGCAGCCCUUGGCCCUGGAGGGGUCCCUGCAGAAGCGUGGCAUUGUGGAACAAUGCUGUACCAGCAUCUGCUCCCUCUACCAGCUGGAGAACUACUGCAACUAG
- example_title: cyclin dependent kinase inhibitor 2A
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AUGGAGCCGGCGGCGGGGAGCAGCAUGGAGCCUUCGGCUGACUGGCUGGCCACGGCCGCGGCCCGGGGUCGGGUAGAGGAGGUGCGGGCGCUGCUGGAGGCGGGGGCGCUGCCCAACGCACCGAAUAGUUACGGUCGGAGGCCGAUCCAGGUCAUGAUGAUGGGCAGCGCCCGAGUGGCGGAGCUGCUGCUGCUCCACGGCGCGGAGCCCAACUGCGCCGACCCCGCCACUCUCACCCGACCCGUGCACGACGCUGCCCGGGAGGGCUUCCUGGACACGCUGGUGGUGCUGCACCGGGCCGGGGCGCGGCUGGACGUGCGCGAUGCCUGGGGCCGUCUGCCCGUGGACCUGGCUGAGGAGCUGGGCCAUCGCGAUGUCGCACGGUACCUGCGCGCGGCUGCGGGGGGCACCAGAGGCAGUAACCAUGCCCGCAUAGAUGCCGCGGAAGGUCCCUCAGACAUCCCCGAUUGA
- example_title: human papillomavirus type 16 E6
pipeline_tag: splice-site
sequence_type: mRNA
task: splice-site
text: AUGCACCAAAAGAGAACUGCAAUGUUUCAGGACCCACAGGAGCGACCCAGAAAGUUACCACAGUUAUGCACAGAGCUGCAAACAACUAUACAUGAUAUAAUAUUAGAAUGUGUGUACUGCAAGCAACAGUUACUGCGACGUGAGGUAUAUGACUUUGCUUUUCGGGAUUUAUGCAUAGUAUAUAGAGAUGGGAAUCCAUAUGCUGUAUGUGAUAAAUGUUUAAAGUUUUAUUCUAAAAUUAGUGAGUAUAGACAUUAUUGUUAUAGUUUGUAUGGAACAACAUUAGAACAGCAAUACAACAAACCGUUGUGUGAUUUGUUAAUUAGGUGUAUUAACUGUCAAAAGCCACUGUGUCCUGAAGAAAAGCAAAGACAUCUGGACAAAAAGCAAAGAUUCCAUAAUAUAAGGGGUCGGUGGACCGGUCGAUGUAUGUCUUGUUGCAGAUCAUCAAGAACACGUAGAGAAACCCAGCUGUAA
- example_title: NRAS proto-oncogene
pipeline_tag: splice-site
sequence_type: 5' UTR
task: splice-site
text: GGGGCCGGAAGUGCCGCUCCUUGGUGGGGGCUGUUCAUGGCGGUUCCGGGGUCUCCAACAUUUUUCCCGGCUGUGGUCCUAAAUCUGUCCAAAGCAGAGGCAGUGGAGCUUGAGGUUCUUGCUGGUGUGAA
- example_title: amyloid beta precursor protein
pipeline_tag: splice-site
sequence_type: 5' UTR
task: splice-site
text: GUCAGUUUCCUCGGCAGCGGUAGGCGAGAGCACGCGGAGGAGCGUGCGCGGGGGCCCCGGGAGACGGCGGCGGUGGCGGCGCGGGCAGAGCAAGGACGCGGCGGAUCCCACUCGCACAGCAGCGCACUCGGUGCCCCGCGCAGGGUCGCG
- example_title: RUNX family transcription factor 1
pipeline_tag: splice-site
sequence_type: 5' UTR
task: splice-site
text: ACUUCUUUGGGCCUCAUAAACAACCACAGAACCACAAGUUGGGUAGCCUGGCAGUGUCAGAAGUCUGAACCCAGCAUAGUGGUCAGCAGGCAGGACGAAUCACACUGAAUGCAAACCACAGGGUUUCGCAGCGUGGUAAAAGAAAUCAUUGAGUCCCCCGCCUUCAGAAGAGGGUGCAUUUUCAGGAGGAAGCG
- example_title: fragile X messenger ribonucleoprotein 1
pipeline_tag: splice-site
sequence_type: 5' UTR
task: splice-site
text: CUCAGUCAGGCGCUCAGCUCCGUUUCGGUUUCACUUCCGGUGGAGGGCCGCCUCUGAGCGGGCGGCGGGCCGACGGCGAGCGCGGGCGGCGGCGGUGACGGAGGCGCCGCUGCCAGGGGGCGUGCGGCAGCGCGGCGGCGGCGGCGGCGGCGGCGGCGGCGGAGGCGGCGGCGGCGGCGGCGGCGGCGGCGGCUGGGCCUCGAGCGCCCGCAGCCCACCUCUCGGGGGCGGGCUCCCGGCGCUAGCAGGGCUGAAGAGAAG
- example_title: MYC proto-oncogene
pipeline_tag: splice-site
sequence_type: 5' UTR
task: splice-site
text: AACUCGCUGUAGUAAUUCCAGCGAGAGGCAGAGGGAGCGAGCGGGCGGCCGGCUAGGGUGGAAGAGCCGGGCGAGCAGAGCUGCGCUGCGGGCGUCCUGGGAAGGGAGAUCCGGAGCGAAUAGGGGGCUUCGCCUCUGGCCCAGCCCUCCCGCUGAUCCCCCAGCCAGCGGUCCGCAACCCUUGCCGCAUCCACGAAACUUUGCCCAUAGCAGCGGGCGGGCACUUUGCACUGGAACUUACAACACCCGAGCAAGGACGCGACUCUCCCGACGCGGGGAGGCUAUUCUGCCCAUUUGGGGACACUUCCCCGCCGCUGCCAGGACCCGCUUCUCUGAAAGGCUCUCCUUGCAGCUGCUUAGACG
- example_title: activating transcription factor 4
pipeline_tag: splice-site
sequence_type: 5' UTR
task: splice-site
text: CAUUUCUACUUUGCCCGCCCACAGAUGUAGUUUUCUCUGCGCGUGUGCGUUUUCCCUCCUCCCCGCCCUCAGGGUCCACGGCCACCAUGGCGUAUUAGGGGCAGCAGUGCCUGCGGCAGCAUUGGCCUUUGCAGCGGCGGCAGCAGCACCAGGCUCUGCAGCGGCAACCCCCAGCGGCUUAAGCCAUGGCGCUUCUCACGGCAUUCAGCAGCAGCGUUGCUGUAACCGACAAAGACACCUUCGAAUUAAGCACAUUCCUCGAUUCCAGCAAAGCACCGCAAC
- example_title: Human GPI protein p137
pipeline_tag: splice-site
sequence_type: 3' UTR
task: splice-site
text: UUUUUAAAAGGAAAAGAUACCAAAUGCCUGCUGCUACCACCCUUUUCAAUUGCUAUGUUUUGAAAGGCACCAGUAUGUGUUUUAGAUUGAUUUAAAUGUUUCAUUUAAAUCACGGACAGUAGUUUCAGUUCUGAUGGUAUAAGCAAAACAAAUAAAACGUUUAUAAAAGUUGUAUCUUGAAACACUGGUGUUCAACAGCUAGCAGCUUAUGUGAUUCACCCCAUGCCACGUUAGUGUCACAAAUUUUAUGGUUUAUCUCCAGCAACAUUUCUCUAGUACUUGCACUUAUUAUCUGAAUUC
- example_title: nucleophosmin 1
pipeline_tag: splice-site
sequence_type: 3' UTR
task: splice-site
text: GAAAAUAGUUUAAACAAUUUGUUAAAAAAUUUUCCGUCUUAUUUCAUUUCUGUAACAGUUGAUAUCUGGCUGUCCUUUUUAUAAUGCAGAGUGAGAACUUUCCCUACCGUGUUUGAUAAAUGUUGUCCAGGUUCUAUUGCCAAGAAUGUGUUGUCCAAAAUGCCUGUUUAGUUUUUAAAGAUGGAACUCCACCCUUUGCUUGGUUUUAAGUAUGUAUGGAAUGUUAUGAUAGGACAUAGUAGUAGCGGUGGUCAGACAUGGAAAUGGUGGGGAGACAAAAAUAUACAUGUGAAAUAAAACUCAGUAUUUUAAUAAAGUAGCACGGUUUCUAUUGA
- example_title: superoxide dismutase 1
pipeline_tag: splice-site
sequence_type: 3' UTR
task: splice-site
text: ACAUUCCCUUGGAUGUAGUCUGAGGCCCCUUAACUCAUCUGUUAUCCUGCUAGCUGUAGAAAUGUAUCCUGAUAAACAUUAAACACUGUAAUCUUAAAAGUGUAAUUGUGUGACUUUUUCAGAGUUGCUUUAAAGUACCUGUAGUGAGAAACUGAUUUAUGAUCACUUGGAAGAUUUGUAUAGUUUUAUAAAACUCAGUUAAAAUGUCUGUUUCAAUGACCUGUAUUUUGCCAGACUUAAAUCACAGAUGGGUAUUAAACUUGUCAGAAUUUCUUUGUCAUUCAAGCCUGUGAAUAAAAACCCUGUAUGGCACUUAUUAUGAGGCUAUUAAAAGAAUCCAAAUUCAAACUAAA
- example_title: hemoglobin subunit alpha 2
pipeline_tag: splice-site
sequence_type: 3' UTR
task: splice-site
text: CUGGAGCCUCGGUAGCCGUUCCUCCUGCCCGCUGGGCCUCCCAACGGGCCCUCCUCCCCUCCUUGCACCGGCCCUUCCUGGUCUUUGAAUAAAGUCUGAGUGGGCAGCA
- example_title: BRAF proto-oncogene
pipeline_tag: splice-site
sequence_type: 3' UTR
task: splice-site
text: AACAAAUGAGUGAGAGAGUUCAGGAGAGUAGCAACAAAAGGAAAAUAAAUGAACAUAUGUUUGCUUAUAUGUUAAAUUGAAUAAAAUACUCUCUUUUUUUUUAAGGUGAACCAAAGAACACUUGUGUGGUUAAAGACUAGAUAUAAUUUUUCCCCAAACUAAAAUUUAUACUUAACAUUGGAUUUUUAACAUCCAAGGGUUAAAAUACAUAGACAUUGCUAAAAAUUGGCAGAGCCUCUUCUAGAGGCUUUACUUUCUGUUCCGGGUUUGUAUCAUUCACUUGGUUAUUUUAAGUAGUAAACUUCAGUUUCUCAUGCAACUUUUGUUGCCAGCUAUCACAUGUCCACUAGGGACUCCAGAAGAAGACCCUACCUAUGCCUGUGUUUGCAGGUGAGAAGUUGGCAGUCGGUUAGCCUGGG
- example_title: H3 clustered histone 1
pipeline_tag: splice-site
sequence_type: 3' UTR
task: splice-site
text: UUACUGUGGUCUCUCUGACGGUCCAAGCAAAGGCUCUUUUCAGAGCCACCACCUUUUC
---
# DeltaSplice
Reference-informed prediction of alternative splicing and splicing-altering mutations from sequences.
## Disclaimer
This is an UNOFFICIAL implementation of [Reference-informed prediction of alternative splicing and splicing-altering mutations from sequences](https://doi.org/10.1101/gr.279044.124) by Chencheng Xu, Suying Bao, et al.
The OFFICIAL repository of DeltaSplice is at [chaolinzhanglab/DeltaSplice](https://github.com/chaolinzhanglab/DeltaSplice).
> [!TIP]
> The MultiMolecule team has confirmed that the provided model and checkpoints are producing the same intermediate representations as the original implementation.
**The team releasing DeltaSplice did not write this model card for this model so this model card has been written by the MultiMolecule team.**
## Model Details
DeltaSplice predicts splice-site usage (SSU) and splicing-altering mutation effects from sequence. The model uses a valid-convolution dilated residual encoder and three prediction modules: splice-site usage, reference-informed delta-SSU, and an auxiliary splice-site head. The official package uses the average prediction of five checkpoints for SSU and delta-SSU prediction; MultiMolecule stores the five seed checkpoints of each released data variant as internal ensemble members and returns their average prediction.
### Variants
- **[multimolecule/deltasplice](https://huggingface.co/multimolecule/deltasplice)**: DeltaSplice trained on the multi-species training set used by the upstream default checkpoints.
- **[multimolecule/deltasplice-human](https://huggingface.co/multimolecule/deltasplice-human)**: DeltaSplice human-only comparison checkpoint set released by the upstream project.
### Model Specification
| Variant | Num Layers | Hidden Size | Context | Ensemble Members | Num Parameters (M) | FLOPs (M) | MACs (M) |
| ----------------- | ---------- | ----------- | ------- | ---------------- | ------------------ | ---------- | --------- |
| DeltaSplice | 24 | 64 | 30000 | 5 | 40.376 | 1642965.72 | 820284.36 |
| DeltaSplice-Human | 24 | 64 | 30000 | 5 | 40.376 | 1642965.72 | 820284.36 |
(FLOPs and MACs measured on one requested output nucleotide with the default 30 kb padded context.)
### Links
- **Code**: [multimolecule.deltasplice](https://github.com/DLS5-Omics/multimolecule/tree/master/multimolecule/models/deltasplice)
- **Paper**: [Reference-informed prediction of alternative splicing and splicing-altering mutations from sequences](https://doi.org/10.1101/gr.279044.124)
- **Developed by**: Chencheng Xu, Suying Bao, Ye Wang, Wenxing Li, Hao Chen, Yufeng Shen, Tao Jiang, Chaolin Zhang
- **Model type**: Dilated residual 1D CNN ensemble for splice-site usage and delta-SSU prediction
- **Original Repository**: [chaolinzhanglab/DeltaSplice](https://github.com/chaolinzhanglab/DeltaSplice)
## Usage
The model file depends on the [`multimolecule`](https://multimolecule.danling.org) library. You can install it using pip:
```bash
pip install multimolecule
```
### Direct Use
#### Splice-Site Usage
```python
>>> from multimolecule import RnaTokenizer
>>> from multimolecule.models.deltasplice import DeltaSpliceModel
>>> tokenizer = RnaTokenizer.from_pretrained("multimolecule/deltasplice-human")
>>> model = DeltaSpliceModel.from_pretrained("multimolecule/deltasplice-human")
>>> inputs = tokenizer("AGCAGUCAUUAUGGCGAAUCUGGCAAGUA", return_tensors="pt")
>>> output = model(**inputs)
>>> output["probabilities"].shape
torch.Size([1, 30, 3])
```
#### Variant Effect
```python
>>> from multimolecule import RnaTokenizer
>>> from multimolecule.models.deltasplice import DeltaSpliceModel
>>> tokenizer = RnaTokenizer.from_pretrained("multimolecule/deltasplice-human")
>>> model = DeltaSpliceModel.from_pretrained("multimolecule/deltasplice-human")
>>> reference = tokenizer("AGCAGUCAUUAUGGCGAAUCUGGCAAGUA", return_tensors="pt")
>>> alternative = tokenizer("AGCAGUCAUUAUGGCUAAUCUGGCAAGUA", return_tensors="pt")
>>> output = model(reference["input_ids"], alternative_input_ids=alternative["input_ids"], use_reference=True)
>>> output["delta"].shape
torch.Size([1, 30, 3])
```
### Interface
- **Input**: RNA sequence tokenized with `RnaTokenizer`; `N` is encoded as zero nucleotide channels
- **Output channels**: `no_splice`, `acceptor`, `donor`
- **Reference-only call**: returns per-position splice-site usage probabilities in `probabilities`
- **Reference + alternative call**: pass the reference sequence as `input_ids` and the alternate sequence as `alternative_input_ids`
- **Reference usage**: pass `reference_usage` with shape `(batch_size, sequence_length, 3)` or omit it to use the model's own reference usage as the reference signal
## Training Details
DeltaSplice was trained to predict splice-site usage from gene sequence and to improve mutation-effect prediction by incorporating reference splice-site usage.
### Training Data
The upstream repository describes training from `gene_dataset.tsu.txt`, which contains splice-site usage in adult brains of eight mammalian species.
### Training Procedure
The official release provides five seed checkpoints with the same architecture and data split. MultiMolecule represents these seed checkpoints as internal ensemble members rather than public model variants.
## Citation
```bibtex
@article{xu2024deltasplice,
title = {Reference-informed prediction of alternative splicing and splicing-altering mutations from sequences},
author = {Xu, Chencheng and Bao, Suying and Wang, Ye and Li, Wenxing and Chen, Hao and Shen, Yufeng and Jiang, Tao and Zhang, Chaolin},
journal = {Genome Research},
volume = {34},
number = {7},
pages = {1052--1065},
year = {2024},
doi = {10.1101/gr.279044.124}
}
```
> [!NOTE]
> The artifacts distributed in this repository are part of the MultiMolecule project.
> If MultiMolecule supports your research, please cite the MultiMolecule project as follows:
```bibtex
@software{chen_2024_12638419,
author = {Chen, Zhiyuan and Zhu, Sophia Y.},
title = {MultiMolecule},
doi = {10.5281/zenodo.12638419},
publisher = {Zenodo},
url = {https://doi.org/10.5281/zenodo.12638419},
year = 2024,
month = may,
day = 4
}
```
## Contact
Please use GitHub issues of [MultiMolecule](https://github.com/DLS5-Omics/multimolecule/issues) for any questions or comments on the model card.
Please contact the authors of the [DeltaSplice paper](https://doi.org/10.1101/gr.279044.124) for questions or comments on the paper/model.
## License
This model implementation is licensed under the [GNU Affero General Public License](license.md).
For additional terms and clarifications, please refer to our [License FAQ](license-faq.md).
```spdx
SPDX-License-Identifier: AGPL-3.0-or-later
``` |