Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,17 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# ESMC Protein Function Predictor
|
| 2 |
|
| 3 |
An Evolutionary-scale Model (ESM) for protein function prediction from amino acid sequences using the Gene Ontology (GO). Based on the ESM Cambrian Transformer architecture, pre-trained on [UniRef](https://www.uniprot.org/help/uniref), [MGnify](https://www.ebi.ac.uk/metagenomics), and the Joint Genome Institute's database and fine-tuned on the [AmiGO Boost](https://huggingface.co/datasets/andrewdalpino/AmiGO-Boost) protein function dataset, this model predicts the GO subgraph for a particular protein sequence - giving you insight into the molecular function, biological process, and location of the activity inside the cell.
|
|
@@ -17,7 +31,7 @@ The following pretrained models are available on HuggingFace Hub.
|
|
| 17 |
| Name | Embedding Dim. | Attn. Heads | Encoder Layers | Context Length | QAT | Total Parameters |
|
| 18 |
|---|---|---|---|---|---|---|
|
| 19 |
| [andrewdalpino/ESMC-300M-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-300M-Protein-Function) | 960 | 15 | 30 | 2048 | None | 361M |
|
| 20 |
-
| [andrewdalpino/ESMC-300M-QAT-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-300M-QAT-Protein-Function) | 960 | 15 | 30 | 2048 |
|
| 21 |
| [andrewdalpino/ESMC-600M-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-600M-Protein-Function) | 1152 | 18 | 36 | 2048 | None | 644M |
|
| 22 |
| [andrewdalpino/ESMC-600M-QAT-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-600M-QAT-Protein-Function) | 1152 | 18 | 36 | 2048 | int8w | 644M |
|
| 23 |
|
|
@@ -74,4 +88,4 @@ The training code can be found at [https://github.com/andrewdalpino/ESMC-Functio
|
|
| 74 |
## References:
|
| 75 |
|
| 76 |
>- T. Hayes, et al. Simulating 500 million years of evolution with a language model, 2024.
|
| 77 |
-
>- M. Ashburner, et al. Gene Ontology: tool for the unification of biology, 2000.
|
|
|
|
| 1 |
+
---
|
| 2 |
+
datasets:
|
| 3 |
+
- andrewdalpino/AmiGO-Boost
|
| 4 |
+
metrics:
|
| 5 |
+
- precision
|
| 6 |
+
- recall
|
| 7 |
+
- f1
|
| 8 |
+
base_model:
|
| 9 |
+
- EvolutionaryScale/esmc-300m-2024-12
|
| 10 |
+
pipeline_tag: text-classification
|
| 11 |
+
tags:
|
| 12 |
+
- gene-ontology
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
# ESMC Protein Function Predictor
|
| 16 |
|
| 17 |
An Evolutionary-scale Model (ESM) for protein function prediction from amino acid sequences using the Gene Ontology (GO). Based on the ESM Cambrian Transformer architecture, pre-trained on [UniRef](https://www.uniprot.org/help/uniref), [MGnify](https://www.ebi.ac.uk/metagenomics), and the Joint Genome Institute's database and fine-tuned on the [AmiGO Boost](https://huggingface.co/datasets/andrewdalpino/AmiGO-Boost) protein function dataset, this model predicts the GO subgraph for a particular protein sequence - giving you insight into the molecular function, biological process, and location of the activity inside the cell.
|
|
|
|
| 31 |
| Name | Embedding Dim. | Attn. Heads | Encoder Layers | Context Length | QAT | Total Parameters |
|
| 32 |
|---|---|---|---|---|---|---|
|
| 33 |
| [andrewdalpino/ESMC-300M-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-300M-Protein-Function) | 960 | 15 | 30 | 2048 | None | 361M |
|
| 34 |
+
| [andrewdalpino/ESMC-300M-QAT-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-300M-QAT-Protein-Function) | 960 | 15 | 30 | 2048 | int8w | 361M |
|
| 35 |
| [andrewdalpino/ESMC-600M-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-600M-Protein-Function) | 1152 | 18 | 36 | 2048 | None | 644M |
|
| 36 |
| [andrewdalpino/ESMC-600M-QAT-Protein-Function](https://huggingface.co/andrewdalpino/ESMC-600M-QAT-Protein-Function) | 1152 | 18 | 36 | 2048 | int8w | 644M |
|
| 37 |
|
|
|
|
| 88 |
## References:
|
| 89 |
|
| 90 |
>- T. Hayes, et al. Simulating 500 million years of evolution with a language model, 2024.
|
| 91 |
+
>- M. Ashburner, et al. Gene Ontology: tool for the unification of biology, 2000.
|