药物发现
药物发现是将机器学习技术应用于新候选药物的识别与开发过程的任务。其目标在于通过计算模型预测化合物活性,优化药物设计流程,提高发现潜在治疗药物的效率和成功率,从而加速药物研发周期,降低研发成本,提升医疗健康领域的创新能力和治疗水平。
QM9
PAMNet
Tox21
elEmBERT-V1
BACE
HIV dataset
GraphConv + dummy super node + focal loss
MUV
GraphConv + dummy super node
ToxCast
BBBP
ProtoW-L2
BindingDB
AttentionSiteDTI
clintox
BiLSTM
DAVIS-DTA
KIBA
SMT-DTA
LIT-PCBA(ALDH1)
LIT-PCBA(KAT2A)
EGT+TGT-At-DP
LIT-PCBA(MAPK1)
SIDER
Ensemble locally constant networks
LIT-PCBA(ESR1_ant)
BindingDB IC50
DeepDTA
PCBA
GraphConv + dummy super node
BACE (β-secretase enzyme)
BBBP (Blood-Brain Barrier Penetration)
DRD2
egfr-inh
Multi-input Neural network with Attention
ESOL (Estimated SOLubility)
FreeSolv (Free Solvation)
Lipophilicity (logd74)
PDBbind
Ensemble locally constant networks
QED
HierG2G
ToxCast (Toxicity Forecaster)
GLAM