Drug Similarity Integration Through Attentive Multi-view Graph Auto-Encoders

Authors: Tengfei Ma, Cao Xiao, Jiayu Zhou, Fei Wang

IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrated significant predictive accuracy improvement. Case studies also showed better model capacity (e.g. embed node features) and interpretability.
Researcher Affiliation Collaboration Tengfei Ma1,2, Cao Xiao1,2, Jiayu Zhou3, Fei Wang4 1 IBM Research 2 MIT-IBM Watson AI Lab 3 Computer Science and Engineering, Michigan State University 4 Weill Cornell Medical School, Cornell University
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper does not provide any explicit statement about releasing source code or a link to a code repository for their methods.
Open Datasets Yes 1) DDI: The known labels of DDIs are extracted from the Twosides database [Tatonetti et al., 2012] (...) 2) Label Side Effect: Drugs side effects extracted from SIDER database [Kuhn et al., 2015] (...) 3) Off-Label Side Effect: Drugs confounder-controlled side effects from OFFSIDES dataset (...) 4) Chemical Structure: Drug structure features (...) using the R package rcdk [?]. (...) 1) Drug Indication: The drug indication data of dimension 1702 is downloaded from SIDER [Kuhn et al., 2015]. (...) 2) Drug chemical protein interactome (CPI): The CPI data from [Wishart et al., 2008] (...) 3) Protein and nucleic acid targets (TTD): (...) extracted from the Therapeutic Target Database (TTD) [Chen et al., 2002].
Dataset Splits Yes In evaluation, we adopted strategies in [Zhang et al., 2015] and randomly selected a fixed percentage (i.e., 25% and 50%) of drugs, and moved all DDIs associated with these drugs for testing. For the data not in testing, we train on 90% and perform validation and model selection on 10% of the drugs.
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments (e.g., GPU/CPU models, memory specifications).
Software Dependencies Yes We implement the proposed model with Tensorflow 1.0 [Abadi et al., 2015] and trained using Adam with learning rate 0.01 and early stopping with window size 30. (...) For MKL, we used the python Mklaren library [Strazar and Curk, 2016].
Experiment Setup Yes We implement the proposed model with Tensorflow 1.0 [Abadi et al., 2015] and trained using Adam with learning rate 0.01 and early stopping with window size 30. We optimized the hyperparameter for Semi GAE on validation data and then fixed for all GAE models: 0.5 (dropout rate), 5e-4 (L2 regularization) and 64 (# of hidden units). For GCN models, we have the second layer and the number of the hidden units in the second layer is set as 32.