Translation
Transformers
PyTorch
t5
text2text-generation
chemistry
biology
text-generation-inference
Instructions to use AI4PD/REXzyme with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use AI4PD/REXzyme with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="AI4PD/REXzyme")# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("AI4PD/REXzyme") model = AutoModelForSeq2SeqLM.from_pretrained("AI4PD/REXzyme") - Notebooks
- Google Colab
- Kaggle
Update README.md
Browse files
README.md
CHANGED
|
@@ -23,20 +23,22 @@ inference:
|
|
| 23 |
# **REXzyme: A Translation Machine for the Generation of New-to-Nature Enzymes**
|
| 24 |
**Work in Progress**
|
| 25 |
|
| 26 |
-
REXzyme (Reaction to Enzyme) (manuscript in preparation) is a translation machine
|
|
|
|
| 27 |
|
| 28 |

|
| 29 |
|
| 30 |
It is possible to provide fine-grained input at the substrate level.
|
| 31 |
Akin to how translation machines have learned to translate between complex language pairs with great success,
|
| 32 |
-
often diverging in their representation at the character level
|
| 33 |
-
be able to translate between the chemical and sequence spaces. REXzyme was trained on a set of 2480 reactions
|
| 34 |
-
sequences that
|
| 35 |
|
| 36 |
-
To run it, you will need to provide a reaction in the SMILE format
|
|
|
|
| 37 |
|
| 38 |
-
After converting each of the reaction components you should combine them in the following scheme
|
| 39 |
-
Additionally
|
| 40 |
e.g. for the carbonic anhydrase ```r2sO.COO>>HCOOO.[H+]</s>```
|
| 41 |
|
| 42 |
or via this simple python script:
|
|
|
|
| 23 |
# **REXzyme: A Translation Machine for the Generation of New-to-Nature Enzymes**
|
| 24 |
**Work in Progress**
|
| 25 |
|
| 26 |
+
REXzyme (Reaction to Enzyme) (manuscript in preparation) is a translation machine, similar to Google Translator,
|
| 27 |
+
for the generation of enzymes that catalize user-defined reactions.
|
| 28 |
|
| 29 |

|
| 30 |
|
| 31 |
It is possible to provide fine-grained input at the substrate level.
|
| 32 |
Akin to how translation machines have learned to translate between complex language pairs with great success,
|
| 33 |
+
often diverging in their representation at the character level (Japanese - English), we posit that an advanced architecture will
|
| 34 |
+
be able to translate between the chemical and sequence spaces. REXzyme was trained on a set of 2480 reactions
|
| 35 |
+
and ~32M enzyme pairs and it produces sequences that are predicted to perform their intended reactions.
|
| 36 |
|
| 37 |
+
To run it, you will need to provide a reaction in the SMILE format
|
| 38 |
+
(Simplified molecular-input line-entry system), which you can do online here: https://cactus.nci.nih.gov/chemical/structure.
|
| 39 |
|
| 40 |
+
After converting each of the reaction components you should combine them in the following scheme: ```ReactantA.ReactantB>AgentA>ProductA.ProductB```<br/>
|
| 41 |
+
Additionally prepending the task suffix ```r2s``` and append the eos token ```</s>```
|
| 42 |
e.g. for the carbonic anhydrase ```r2sO.COO>>HCOOO.[H+]</s>```
|
| 43 |
|
| 44 |
or via this simple python script:
|