0.25
0.5
0.75
1.25
1.5
1.75
2
Kernel-based predictive modelling of drug-protein binding affinities
Published on Jun 28, 201971 Views
Despite several years of target-based drug discovery, chemical agents inhibiting single targets are still rare. Mapping the complete protein target space of drugs and drug-like compounds, including bo
Related categories
Chapter list
Kernel-based predictive modelling of drug–protein binding affinities00:00
Drug–protein interaction - 100:34
Drug–protein interaction - 201:00
Off-target interactions01:19
Drug–protein interaction profiling02:13
Enormous chemical universe02:37
Motivation for machine learning in drug discovery03:17
Computational drug screening03:48
Systems-based DPI prediction methods05:09
Similarities between molecules06:31
Kernels - 106:47
Kernels - 208:35
Kernels - 309:25
Kernels - 410:19
New DPIs prediction - 111:54
Known interactions – bioactivity databases12:10
New DPIs prediction - 213:55
Molecular fingerprint14:01
2D fingerprints14:46
3D fingerprints16:02
Fingerprint-based Tanimoto kernel16:23
Graph kernel17:09
Graph 18:07
Random walk18:30
Random walk graph kernel19:00
Problem with using chemical structures21:06
Additional information21:47
New DPIs prediction - 323:33
Amino acid sequence alignment23:37
Generic String (GS) kernel - 124:11
Generic String (GS) kernel - 227:40
Generic String (GS) kernel - 327:53
Additional information28:05
New DPIs prediction - 429:22
Supervised learning29:37
Overfitting - 130:23
Overfitting - 230:32
Overfitting - 330:42
Overfitting - 430:50
Regularization30:58
Linear model31:31
Ridge regression - 132:25
Linear model and dual representation33:25
Ridge regression - 234:57
Kernel ridge regression (KRR)35:15
KRR for drug-protein binding affinity prediction35:38
Pairwise kernel36:17
Kronecker product36:29
Pairwise kernel36:51
Pairwise KRR – shortcut37:50
Multiple Kernel Learning (MKL)41:03
Two-stage MKL - 144:10
Two-stage MKL - 244:54
Two-stage MKL - 352:48
MKL with pairwise kernels53:46
pairwiseMKL - 154:12
pairwiseMKL - 254:39
pairwiseMKL - 354:54
pairwiseMKL - 457:25
pairwiseMKL - 559:23
pairwiseMKL - 601:00:25
pairwiseMKL - 701:00:58
pairwiseMKL: application example01:01:13
Drug kernels01:01:59
Protein kinase kernels01:02:07
pairwiseMKL: CV results01:04:22
Pairwise prediction scenarios01:07:26
Overall aims01:09:51
Scope01:10:52
Overview of the IDG-DREAM Drug-Kinase Binding Prediction Challenge01:11:19
Results: Round 1 vs. Round 201:14:00
Thank you!01:15:57