Views
No views yet
pip install deepchemography1from deepchemography.peptides import load_peptide_model, sample_peptides
2
3# Load model
4model, vocab = load_peptide_model("path/to/model.pt", "path/to/vocab.dict")
5
6# Sample new peptides
7peptides = sample_peptides(model, vocab, n_samples=10)
8for p in peptides:
9 print(p) # e.g., "M L L L L L A L A L L A L L L"1from deepchemography.peptides import encode_peptide, decode_latent
2
3# Encode a peptide to latent space
4sequence = "M L L L L L A L A L L A L L L A L L L"
5z = encode_peptide(model, vocab, sequence)
6print(f"Latent shape: {z.shape}") # (1, 100)
7
8# Decode latent vectors back to sequences
9reconstructed = decode_latent(model, vocab, z, sample_mode='greedy')
10print(f"Reconstructed: {reconstructed[0]}")1from deepchemography.peptides import interpolate_peptides
2
3seq1 = "M L L L L L A L A L L A L L L A L L L"
4seq2 = "M D K L I V L K M L N S K L P Y G Q R K"
5
6# Linear interpolation in latent space
7sequences, weights = interpolate_peptides(
8 model, vocab, seq1, seq2,
9 n_steps=5,
10 method='linear'
11)
12
13for w, seq in zip(weights, sequences):
14 print(f"w={w:.2f}: {seq}")1from deepchemography.peptides import explore_neighborhood
2
3base_sequence = "M L L L L L A L A L L A L L L A L L L"
4
5# Generate similar peptides
6neighbors = explore_neighborhood(
7 model, vocab, base_sequence,
8 noise_scale=0.1, # Low noise = high similarity
9 n_neighbors=10
10)
11
12for neighbor in neighbors:
13 print(neighbor)M L L L L L A L A L L A L L L A L L LA, C, D, E, F, G, H, I, K, L, M, N, P, Q, R, S, T, U, V, W, Y, Z<unk> (index 0): Unknown token<pad> (index 1): Padding<start> (index 2): Start of sequence<eos> (index 3): End of sequence1@software{wae_peptides,
2 title={Peptide Wasserstein Autoencoder},
3 author={Orlov, Alexander},
4 year={2025},
5 url={https://huggingface.co/axelrolov/wae_peptides}
6}