File size: 5,090 Bytes
6f1e670
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
150
151
152
153
154
155
156
157
158
159
160
161
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Introduction\n",
    "\n",
    "This notebook will discuss how to use the various featurizers in the ```multievolve``` package."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "from model.splitters import *\n",
    "from model.featurizers import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Setting up\n",
    "\n",
    "First, define the following variables:\n",
    "\n",
    "- ```protein_name```: the name of the protein\n",
    "\n",
    "- ```wt_file```: the path to the wildtype sequence\n",
    "\n",
    "- ```training_dataset_fname```: the path to the training dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "protein_name = \"example_protein\"\n",
    "wt_file = \"../../../data/example_protein/apex.fasta\"\n",
    "training_dataset_fname = '../../../data/example_protein/example_dataset.csv'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define a splitter object 鈥?we will just use this to load the dataset and pull sequences from to featurizer later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "splitter = RandomProteinSplitter(protein_name, training_dataset_fname, wt_file, csv_has_header=True, use_cache=True, y_scaling=False, val_split=None)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Featurizers\n",
    "\n",
    "There are many featurizers available in the ```multievolve``` package. We discuss a few of the most common ones below. \n",
    "\n",
    "- ```OneHotFeaturizer```: One-hot encoding of the protein sequence\n",
    "\n",
    "- ```GeorgievFeaturizer```: Georgiev et al. (2022) featurizer\n",
    "\n",
    "- ```AAIdxFeaturizer```: amino acid index featurizer    \n",
    "\n",
    "- ```ESMLogitsFeaturizer```: ESM-2 logits featurizer\n",
    "\n",
    "- ```ESM2EmbedFeaturizer```: ESM-2 embedding featurizer\n",
    "\n",
    "There are also combinatorial featurizers that combine multiple featurizers.\n",
    "\n",
    "- ```ESMAugmentedFeaturizer```: One-hot encoding augmented likelihood scores from the ESM-1/ESM-2 models\n",
    "\n",
    "- ```OnehotAndGeorgievFeaturizer```: One-hot encoding combined with Georgiev et al. (2022) featurizer, wherein the encodings are stacked along the last axis (i.e. by position)\n",
    "\n",
    "- ```OnehotAndAAIdxFeaturizer```: One-hot encoding augmented with amino acid index featurizer, wherein the encodings are stacked along the last axis (i.e. by position)\n",
    "\n",
    "- ```OnehotAndESMLogitsFeaturizer```: One-hot encoding augmented with ESM-2 logits featurizer, wherein the encodings are stacked along the last axis (i.e. by position)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Base Featurizers\n",
    "onehot = OneHotFeaturizer(protein=protein_name, use_cache=True)\n",
    "georgiev = GeorgievFeaturizer(protein=protein_name, use_cache=True)\n",
    "aa_idx = AAIdxFeaturizer(protein=protein_name, use_cache=True)\n",
    "esm_logits = ESMLogitsFeaturizer(protein=protein_name, use_cache=True)\n",
    "esm_embed = ESM2EmbedFeaturizer(protein=protein_name, use_cache=True)\n",
    "\n",
    "# Combinatorial Featurizers\n",
    "esm_augmented = ESMAugmentedFeaturizer(protein=protein_name, use_cache=True, wt_file=wt_file)\n",
    "onehotgeorgiev = OnehotAndGeorgievFeaturizer(protein=protein_name, use_cache=True)\n",
    "onehotaaidx = OnehotAndAAIdxFeaturizer(protein=protein_name, use_cache=True)\n",
    "onehotesmlogits = OnehotAndESMLogitsFeaturizer(protein=protein_name, use_cache=True)\n",
    "onehotesmmsalogits = OnehotAndESMMSALogitsFeaturizer(protein=protein_name, use_cache=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Featurizers have the function ```featurize```, which takes in a list of sequences and returns the featurized sequences."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "example_sequences = splitter.data[0][:5].tolist()\n",
    "\n",
    "onehot.featurize(example_sequences)"
   ]
  }
 ],
 "metadata": {
  "environment": {
   "kernel": "multievolve",
   "name": "workbench-notebooks.m128",
   "type": "gcloud",
   "uri": "us-docker.pkg.dev/deeplearning-platform-release/gcr.io/workbench-notebooks:m128"
  },
  "kernelspec": {
   "display_name": "multievolve",
   "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": "3.11.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}