Network embeddings for modeling polypharmacy and drug-drug interactions thumbnail
Pause
Mute
Subtitles not available
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Network embeddings for modeling polypharmacy and drug-drug interactions

Published on Jun 28, 2019117 Views

Polypharmacy, the use of drug combinations, is common to treat patients with complex or co-existing diseases. However, a major consequence of polypharmacy is a high risk of adverse side effects, which

Related categories

Chapter list

Part 2: Next-Gen Machine Learning for Network Biology00:00
Two Lectures00:11
Networks allow for integration of biomedical data01:24
Outline of this Lecture - 101:57
Deep Learning for Multimodal Networks03:23
Embedding Nodes - 103:33
Embedding Nodes - 204:06
Two Key Components04:32
So Far: Shallow Encoders04:44
Shallow Encoders05:51
Deep Graph Encoders07:11
A Naïve Approach07:45
Approach: Deep Learning for Multimodal Networks11:44
Multimodal Networks12:01
Preview of Results14:03
Overview of our deep learning approach for multimodal networks14:17
Embedding Nodes - 314:51
Key Idea: Aggregate Neighbors15:46
Example: Aggregate Neighbors17:12
Every node learns how to aggregate its own neighbors18:46
Deep Model: Many Layers19:26
The Math: Deep Graph Encoder21:59
Overview of our deep learning approach for multimodal networks - 125:48
Heterogeneous Edge Decoder27:15
Overview of our deep learning approach for multimodal networks - 231:00
Recap: Deep Learning for Multimodal Networks31:58
Outline of this Lecture - 239:29
Polypharmacy and Drug-Drug Interactions39:40
Polypharmacy39:42
Reports on Unwanted Side Effects - 141:09
Reports on Unwanted Side Effects - 242:38
Unexpected Drug Interactions43:59
Why is modeling polypharmacy hard?45:02
We need Polypharmacy Dataset47:26
We apply our deep approach to the polypharmacy network49:37
Results: Side Effect Prediction51:19
Novel Predictions53:20
Utility of Predictions in the Clinic54:52
Validation in the Clinic: Key Idea55:47
Outline of this Lecture - 359:58
Computational Drug Repurposing01:00:10
Computational Drug Discovery01:00:20
New tricks for old drugs01:00:51
Key Insight: Subgraphs01:01:37
Link Prediction Between Subgraphs01:03:18
SUGAR: Neural Message Passing01:04:42
We need Drug Repurposing Dataset01:06:14
Predictive Performance01:07:58
Side Information further improves performance01:09:35
Drug Repurposing at Stanford01:10:52
Introducing Feedbacks for AI Loop01:13:34
Outline of this Lecture - 301:14:14
New Directions and Opportunities01:14:37
New Directions01:14:50
1st New Direction: Explanations01:14:58
2nd New Direction: Train with Less Data01:15:47
3rd New Direction: Rich Interactions01:18:15
Outline of this Lecture - 401:20:51
Practical Advice and Demos01:20:55
Deep Learning for Network Biology01:20:57
Lecture Resources01:21:02
Easy Deep Learning on Graphs01:21:43
Recap01:22:26
Network Biology and Medicine01:25:04
How can this technology be used for biomedical problems?01:25:20