Skip to content

Commit

Permalink
newer updates
Browse files Browse the repository at this point in the history
  • Loading branch information
@SaniyaKhullar
SaniyaKhullar committed Oct 23, 2023
1 parent 56b76ce commit 7ab6453
Showing 1 changed file with 148 additions and 21 deletions.
169 changes: 148 additions & 21 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -273,7 +273,7 @@ import netrem_evaluation_functions as nm_eval

dummy_data = generate_dummy_data(corrVals = [0.9, 0.5, 0.4, -0.3, -0.8], # the # of elements in corrVals is the # of predictors (X)
num_samples_M = 100000, # the number of samples M
train_data_percent = 70) # the remainder out of 100 will be kept for testing. If 100, all data is used for training and testing.
train_data_percent = 70) # the remainder out of 100,000 will be kept for testing. If 100, then ALL data is used for training and testing.
```
The Python console or Jupyter notebook will print out the following:

Expand Down Expand Up @@ -403,21 +403,161 @@ y_test = dummy_data.view_y_test_df()
Our generated single-cell gene expression data (X, y) looks like this:
<img src="netrem_expression_demo.png" style="width: 200px;"/>

```python
X_train.corr() # pairwise correlations among the training samples
```

<div>
<table border="1" class="dataframe">
<thead>
<tr style="text-align: right;">
<th></th>
<th>TF1</th>
<th>TF2</th>
<th>TF3</th>
<th>TF4</th>
<th>TF5</th>
</tr>
</thead>
<tbody>
<tr>
<th>TF1</th>
<td>1.000000</td>
<td>0.444851</td>
<td>0.360777</td>
<td>-0.274352</td>
<td>-0.719915</td>
</tr>
<tr>
<th>TF2</th>
<td>0.444851</td>
<td>1.000000</td>
<td>0.195290</td>
<td>-0.152394</td>
<td>-0.398306</td>
</tr>
<tr>
<th>TF3</th>
<td>0.360777</td>
<td>0.195290</td>
<td>1.000000</td>
<td>-0.125259</td>
<td>-0.320436</td>
</tr>
<tr>
<th>TF4</th>
<td>-0.274352</td>
<td>-0.152394</td>
<td>-0.125259</td>
<td>1.000000</td>
<td>0.242985</td>
</tr>
<tr>
<th>TF5</th>
<td>-0.719915</td>
<td>-0.398306</td>
<td>-0.320436</td>
<td>0.242985</td>
<td>1.000000</td>
</tr>
</tbody>
</table>
</div>




```python
X_test.corr() # pairwise correlations among the testing samples
```




<div>
<style scoped>
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}

.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}
</style>
<table border="1" class="dataframe">
<thead>
<tr style="text-align: right;">
<th></th>
<th>TF1</th>
<th>TF2</th>
<th>TF3</th>
<th>TF4</th>
<th>TF5</th>
</tr>
</thead>
<tbody>
<tr>
<th>TF1</th>
<td>1.000000</td>
<td>0.440370</td>
<td>0.359511</td>
<td>-0.269646</td>
<td>-0.721168</td>
</tr>
<tr>
<th>TF2</th>
<td>0.440370</td>
<td>1.000000</td>
<td>0.189575</td>
<td>-0.146711</td>
<td>-0.389451</td>
</tr>
<tr>
<th>TF3</th>
<td>0.359511</td>
<td>0.189575</td>
<td>1.000000</td>
<td>-0.119603</td>
<td>-0.312932</td>
</tr>
<tr>
<th>TF4</th>
<td>-0.269646</td>
<td>-0.146711</td>
<td>-0.119603</td>
<td>1.000000</td>
<td>0.241384</td>
</tr>
<tr>
<th>TF5</th>
<td>-0.721168</td>
<td>-0.389451</td>
<td>-0.312932</td>
<td>0.241384</td>
<td>1.000000</td>
</tr>
</tbody>
</table>
</div>

```python
# prior network edge_list (missing edges or edges with no edge weight will be added with the default_edge_list so the network is fully-connected):
edge_list = [["TF1", "TF2", 0.9], ["TF4", "TF5", 0.75], ["TF1", "TF3"], ["TF1", "TF4"], ["TF1", "TF5"],
edge_list = [["TF1", "TF2", 0.8], ["TF4", "TF5", 0.95], ["TF1", "TF3"], ["TF1", "TF4"], ["TF1", "TF5"],
["TF2", "TF3"], ["TF2", "TF4"], ["TF2", "TF5"], ["TF3", "TF4"], ["TF3", "TF5"]]

beta_network_val = 3
# by default, model_type is Lasso, so alpha_lasso_val will be specified for the alpha_lasso parameter. (Otherwise, if model_type is LassoCV, alpha_lasso is determined by cross-validation on training data).
alpha_lasso_val = 0.01
beta_network_val = 1
# by default, model_type is Lasso, so alpha_lasso_val will be specified for the alpha_lasso parameter.
# However, we will specify model_type = LassoCV, so our alpha_lasso is determined by cross-validation on training data).

# Building the network regularized regression model:
# Please note: To include nodes found in the gene expression data that are not found in the PPI Network (e.g. TF6 in our case), we use False for the overlapped_nodes_only argument (otherwise, we would only use TFs 1 to 5).
# By default, edges are constructed between all of the nodes; nodes with a missing edge are assigned the default_edge_weight.
netrem_demo = netrem(edge_list = edge_list,
beta_net = beta_network_val,
alpha_lasso = alpha_lasso_val,
model_type = "LassoCV",
view_network = True)

# Fitting the NetREm model on training data: X_train and y_train:
Expand All @@ -428,25 +568,12 @@ netrem_demo.fit(X_train, y_train)
![png](output_3_1.png)





![png](output_3_2.png)



1 new node(s) added to network based on gene expression data ['TF6']



![png](output_3_5.png)

![png](netrem_estimator.PNG)

<!-- There is a particularly strong relationship between $TF_{1} \leftrightarrow TF_{2}$ of 0.9 and between $TF_{4} \leftrightarrow TF_{5}$ of 0.75. The remaining relationships among the other TFs is assumed to be the default (edge weight of 0.1). -->
<!-- Here, $gregulnet_{demo}$ is an object of the *GRegulNet* class. We fit a model using $X_{train}$ and $y_{train}$ data (70 samples). -->

To view and extract the predicted model coefficients for the predictors:
To view and extract the predicted model coefficients *c* for the predictors:

<!-- ```python
gregulnet_demo.coef
Expand Down

0 comments on commit 7ab6453

Please sign in to comment.