DeepSide - A Deep Learning Framework for Drug Side Effect Prediction
-
A. Ercument Cicek
|
Halil Ibrahim Kuru
|
Oznur Tastan
-
Computational Biology Department, Carnegie Mellon University, Pittsburgh, PA 15213
|
Department of Computer Engineering Bilkent University Ankara Turkey 06800
|
Faculty of Engineering and Natural Sciences, Sabanci University, Istanbul, Turkey 34956
Drug failures due to unforeseen adverse effects at clinical trials pose health risks for the participants and lead to substantial financial losses. Side effect prediction algorithms have the potential to guide the drug design process.--This DeepSide framework, use context-related (gene expression) features along with the chemical structure to predict ADRs to account for conditions such as dosing, time interval, and cell line. The proposed MMNN model uses GEX and CS as combined features and achieves better accuracy performance compared to the models that only use the chemical structure (CS) fingerprints. The reported accuracy is noteworthy considering that it is only trying to estimate the condition-independent side effects. Finally, SMILESConv model outperforms all other approaches by applying convolution on SMILES representation of drug chemical structure.
Input variables : Drug checmical structure , Gene Expressions
Output Variables : Binary output for drug realted side effect
Metrics to Monitor
|
Statistical
|
:
|
Somers D |
Accuracy |
Precision and Recall |
Confusion Matrix |
F1 Score |
Roc and Auc |
Prevalence |
Detection Rate |
Balanced Accuracy |
Cohen's Kappa |
Concordance |
Gini Coefficent |
KS Statistic |
Youden's J Index
|
|
Infrastructure
|
:
|
Log Bytes |
Logging/User/IAMPolicy |
Logging/User/VPN |
CPU Utilization |
Memory Usage |
Error Count |
Prediction Count |
Prediction Latencies |
Private Endpoint Prediction Latencies |
Private Endpoint Response Count
|
Visit Model :
researchgate.net