File size: 3,936 Bytes
ca9a9bd 08fff2c ca9a9bd 08fff2c ca9a9bd 08fff2c d3654a6 ca9a9bd 08fff2c ca9a9bd 08fff2c ca9a9bd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 |
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from typing import List, Union\n",
"\n",
"import torch\n",
"from transformers import AutoModel"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Load model"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"model = AutoModel.from_pretrained(\"InstaDeepAI/segment_borzoi\", trust_remote_code=True)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Define useful functions"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"def encode_sequences(sequences: Union[str, List[str]]) -> torch.Tensor:\n",
" \"\"\"\n",
" One-hot encode a DNA sequence or a batch of DNA sequences.\n",
"\n",
" Args:\n",
" sequences (Union[str, List[str]]): Either a DNA sequence or a list of DNA sequences\n",
"\n",
" Returns:\n",
" torch.Tensor: One-hot encoded\n",
" - If `sequences` is just one sequence (str), output shape is (seq_len, 4), seq_len being the length of a sequence\n",
" - If `sequences` is a list of sequences, output shape is (num_sequences, seq_len, 4)\n",
" \n",
" Example:\n",
" >>> sequences = [\"AC\", \"GT\"]\n",
" >>> encode_sequences(sequences)\n",
" tensor([[[1., 0., 0., 0.],\n",
" [0., 1., 0., 0.]],\n",
"\n",
" [[0., 0., 1., 0.],\n",
" [0., 0., 0., 1.]]])\n",
" \"\"\"\n",
" one_hot_map = {\n",
" 'a': torch.tensor([1., 0., 0., 0.]),\n",
" 'c': torch.tensor([0., 1., 0., 0.]),\n",
" 'g': torch.tensor([0., 0., 1., 0.]),\n",
" 't': torch.tensor([0., 0., 0., 1.]),\n",
" 'n': torch.tensor([0., 0., 0., 0.]),\n",
" 'A': torch.tensor([1., 0., 0., 0.]),\n",
" 'C': torch.tensor([0., 1., 0., 0.]),\n",
" 'G': torch.tensor([0., 0., 1., 0.]),\n",
" 'T': torch.tensor([0., 0., 0., 1.]),\n",
" 'N': torch.tensor([0., 0., 0., 0.])\n",
" }\n",
"\n",
" def encode_sequence(seq_str):\n",
" one_hot_list = []\n",
" for char in seq_str:\n",
" one_hot_vector = one_hot_map.get(char, torch.tensor([0.25, 0.25, 0.25, 0.25]))\n",
" one_hot_list.append(one_hot_vector)\n",
" return torch.stack(one_hot_list)\n",
"\n",
" if isinstance(sequences, list):\n",
" return torch.stack([encode_sequence(seq) for seq in sequences])\n",
" else:\n",
" return encode_sequence(sequences)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Inference example"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"sequences = [\"A\"*524_288, \"G\"*524_288]\n",
"one_hot_encoding = encode_sequences(sequences)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"preds = model(one_hot_encoding)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"print(preds['logits'])"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "genomics-research-env",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.10"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|