Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,3 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
import os
|
| 2 |
import gradio as gr
|
| 3 |
|
|
@@ -84,6 +94,8 @@ def sample_at_length(l:int, seed:int):
|
|
| 84 |
|
| 85 |
interface = gr.Interface(
|
| 86 |
fn=sample_at_length,
|
|
|
|
|
|
|
| 87 |
inputs=[
|
| 88 |
gr.Number(value=80, label="Protein backbone length to generate", show_label=True, precision=0),
|
| 89 |
gr.Number(value=42, label="Random seed", show_label=True, precision=0),
|
|
|
|
| 1 |
+
"""
|
| 2 |
+
foldingdiff implements a diffusion model for generating protein structures. Inspired by the biological folding process,
|
| 3 |
+
we perform diffusion on the angles between amino acid residues rather than the absolute 3D coordinates of each residue.
|
| 4 |
+
By effectively treating each residue as its own reference frame, we shift the equivariance constraints into the
|
| 5 |
+
representation space itself; this allows us to use a vanilla transformer model as our model. Here, we provide a simple
|
| 6 |
+
online interface for generating single backbones with a given length, starting from a given random seed.
|
| 7 |
+
|
| 8 |
+
See our preprint at https://arxiv.org/abs/2209.15611 and our full codebase at https://github.com/microsoft/foldingdiff
|
| 9 |
+
"""
|
| 10 |
+
|
| 11 |
import os
|
| 12 |
import gradio as gr
|
| 13 |
|
|
|
|
| 94 |
|
| 95 |
interface = gr.Interface(
|
| 96 |
fn=sample_at_length,
|
| 97 |
+
title="foldingdiff - protein backbone structure generation with diffusion models",
|
| 98 |
+
description=__doc__,
|
| 99 |
inputs=[
|
| 100 |
gr.Number(value=80, label="Protein backbone length to generate", show_label=True, precision=0),
|
| 101 |
gr.Number(value=42, label="Random seed", show_label=True, precision=0),
|