examples/boltz-1.md
... ...
@@ -0,0 +1,64 @@
1
+## Boltz-1
2
+
3
+Boltz-1 is a diffusion-based deep learning model for predicting 3D structures of biomolecular complexes. It can be run on both CPU and GPU systems.
4
+
5
+## Basic Usage
6
+
7
+Boltz-1 can be run with minimal parameters as shown in the examples below.
8
+
9
+### Create dataset
10
+```
11
+mkdir -p test_data
12
+
13
+# Create a minimal MSA file
14
+cat > test_data/test.a3m << EOF
15
+>101
16
+MVKG
17
+>UniRef100_test
18
+MVKG
19
+EOF
20
+
21
+# Create a simple test FASTA with MSA reference
22
+cat > test_data/simple_test.fasta << EOF
23
+>A|protein|./test_data/test.a3m
24
+MVKG
25
+EOF
26
+```
27
+
28
+### Test CPU
29
+
30
+```
31
+# Test with minimal parameters and CPU
32
+boltz predict test_data/simple_test.fasta \
33
+ --accelerator cpu \
34
+ --sampling_steps 25 \
35
+ --recycling_steps 1 \
36
+ --diffusion_samples 1 \
37
+ --cache ./boltz_cache
38
+```
39
+
40
+### Test GPU
41
+**Note:** This test requires a CUDA-capable GPU. If no GPU is available, use the CPU version instead.
42
+
43
+```
44
+# Test with minimal parameters and GPU
45
+boltz predict test_data/simple_test.fasta \
46
+ --accelerator gpu \
47
+ --sampling_steps 25 \
48
+ --recycling_steps 1 \
49
+ --diffusion_samples 1 \
50
+ --cache ./boltz_cache
51
+```
52
+
53
+### Using Your Own Data
54
+
55
+Example with existing data (change `path/to/your_protein.fasta` to your data):
56
+```bash
57
+# Run prediction on your existing FASTA file
58
+boltz predict path/to/your_protein.fasta \
59
+ --accelerator gpu \
60
+ --sampling_steps 25 \
61
+ --recycling_steps 1 \
62
+ --diffusion_samples 1 \
63
+ --cache ./boltz_cache
64
+```