|
11 | 11 | "cell_type": "markdown", |
12 | 12 | "metadata": {}, |
13 | 13 | "source": [ |
14 | | - "**The notebooks in this directory were developed to demonstrate the \"Ten Rules for Reproducible Research with Jupyter Notebooks\". Throughout the notebooks we mention the rules we applied.**\n", |
| 14 | + "**The notebooks in this directory were developed to demonstrate the \"Ten Rules for Reproducible Research with Jupyter Notebooks\". Throughout the notebooks we refer to some the rules we applied.**\n", |
15 | 15 | "\n", |
16 | 16 | "**For example, this notebook demonstrates:**\n", |
17 | 17 | "\n", |
18 | 18 | "---\n", |
19 | 19 | "\n", |
20 | 20 | "**Rule 1: Tell a Story for a Specific Audience.** This notebook was developed for biologists to learn how to apply a simple machine learning model to protein sequences.\n", |
21 | 21 | "\n", |
22 | | - "**Rule 3: Document the Entire Workflow.** This top-level notebook links to 3 notebooks that represent the steps of a workflow. This modularity makes it easy to replace one of the steps, for example, use a different method to calculate features or apply a different machine learning model.\n", |
| 22 | + "**Rule 3: Document the Entire Workflow.** This top-level notebook links to 4 notebooks that represent the steps of a workflow. This modularity makes it easy to replace one of the steps, for example, use a different method to calculate features or apply a different machine learning model.\n", |
23 | 23 | "\n", |
24 | 24 | "---" |
25 | 25 | ] |
|
47 | 47 | "We can classify proteins into three major fold types based on their predominant secondary structure content\n", |
48 | 48 | "* alpha: contains predominantly alpha helices\n", |
49 | 49 | "* beta: contains predominantly beta sheets\n", |
50 | | - "* alpha+beta: contains both alpha helices and beta sheets" |
| 50 | + "* alpha+beta: contains alpha helices and beta sheets" |
51 | 51 | ] |
52 | 52 | }, |
53 | 53 | { |
54 | 54 | "cell_type": "markdown", |
55 | 55 | "metadata": {}, |
56 | 56 | "source": [ |
57 | 57 | "## Goal\n", |
58 | | - "This notebook serves as an example of using machine learning techniques applied to protein sequences. The goal is to create a simple machine learning model to predict the fold type of a protein given its protein sequence. We train the model on a representative set of 3D structure from the Protein Data Bank.\n", |
| 58 | + "This notebook demostrates how to create a reproducible record to create a machine learning model. We train a simple model to predict the fold class of a protein given its protein sequence using a representative set of 3D structures from the Protein Data Bank.\n", |
59 | 59 | "\n", |
60 | 60 | "Run the following notebooks to work through this example." |
61 | 61 | ] |
|
71 | 71 | "cell_type": "markdown", |
72 | 72 | "metadata": {}, |
73 | 73 | "source": [ |
74 | | - "First, we need to create a dataset with protein secondary structure information obtained from 3D protein chains.\n", |
| 74 | + "First, we create a dataset with protein secondary structure information obtained from 3D protein chains.\n", |
75 | 75 | "\n", |
76 | 76 | "Run the following notebook to extract secondary structure information from a representative set of protein chains downloaded from the RCSB Protein Data Bank and assign a fold type to each protein chain." |
77 | 77 | ] |
|
87 | 87 | "cell_type": "markdown", |
88 | 88 | "metadata": {}, |
89 | 89 | "source": [ |
90 | | - "The notebook saves the dataset in the file `secondaryStructure.json`." |
| 90 | + "The notebook saves the dataset in the file `./intermediate_data/foldClassification.json`." |
91 | 91 | ] |
92 | 92 | }, |
93 | 93 | { |
|
117 | 117 | "cell_type": "markdown", |
118 | 118 | "metadata": {}, |
119 | 119 | "source": [ |
120 | | - "The notebook saves the dateset in the file `features.json`." |
| 120 | + "This notebook saves the dataset with feature vectors in the file `./intermediate_data/features.json`." |
121 | 121 | ] |
122 | 122 | }, |
123 | 123 | { |
|
131 | 131 | "cell_type": "markdown", |
132 | 132 | "metadata": {}, |
133 | 133 | "source": [ |
134 | | - "Next, we fit a 3-state classification model using the feature vectors as inputs and the known fold types from the Protein Data Bank dataset.\n", |
| 134 | + "Next, we fit a 3-state classification model using the feature vectors and the given fold classification from the Protein Data Bank dataset.\n", |
135 | 135 | "\n", |
136 | 136 | "Run the following notebook to fit a machine learning model on a training set and evaluate its performance on a test set." |
137 | 137 | ] |
|
143 | 143 | "[3-FitModel.ipynb](./3-FitModel.ipynb)" |
144 | 144 | ] |
145 | 145 | }, |
| 146 | + { |
| 147 | + "cell_type": "markdown", |
| 148 | + "metadata": {}, |
| 149 | + "source": [ |
| 150 | + "This notebook saves the classification model in the file `./intermediate_data/classifier`." |
| 151 | + ] |
| 152 | + }, |
146 | 153 | { |
147 | 154 | "cell_type": "markdown", |
148 | 155 | "metadata": {}, |
|
154 | 161 | "cell_type": "markdown", |
155 | 162 | "metadata": {}, |
156 | 163 | "source": [ |
157 | | - "Finally, we use the Word2Vec model and the trained classifier to predict the fold class from a protein sequence." |
| 164 | + "Finally, we use the trained classifier to predict the fold class from a protein sequence." |
158 | 165 | ] |
159 | 166 | }, |
160 | 167 | { |
|
184 | 191 | }, |
185 | 192 | { |
186 | 193 | "cell_type": "code", |
187 | | - "execution_count": 2, |
| 194 | + "execution_count": 1, |
188 | 195 | "metadata": {}, |
189 | 196 | "outputs": [ |
190 | 197 | { |
191 | 198 | "name": "stdout", |
192 | 199 | "output_type": "stream", |
193 | 200 | "text": [ |
194 | | - "The watermark extension is already loaded. To reload it, use:\n", |
195 | | - " %reload_ext watermark\n", |
196 | 201 | "CPython 3.6.3\n", |
197 | 202 | "IPython 6.3.1\n", |
198 | 203 | "\n", |
199 | | - "gensim 3.6.0\n", |
| 204 | + "ipywidgets 7.4.0\n", |
200 | 205 | "matplotlib 2.2.2\n", |
201 | 206 | "numpy 1.14.5\n", |
202 | 207 | "pandas 0.22.0\n", |
|
214 | 219 | ], |
215 | 220 | "source": [ |
216 | 221 | "%load_ext watermark\n", |
217 | | - "%watermark -v -m -p gensim,matplotlib,numpy,pandas,sklearn" |
| 222 | + "%watermark -v -m -p ipywidgets,matplotlib,numpy,pandas,sklearn" |
| 223 | + ] |
| 224 | + }, |
| 225 | + { |
| 226 | + "cell_type": "markdown", |
| 227 | + "metadata": {}, |
| 228 | + "source": [ |
| 229 | + "---\n", |
| 230 | + "\n", |
| 231 | + "**Authors:** Peter W. Rose, Shih-Cheng Huang, UC San Diego, October 1, 2018\n", |
| 232 | + "\n", |
| 233 | + "---" |
218 | 234 | ] |
219 | 235 | } |
220 | 236 | ], |
|
0 commit comments