en
0.25
0.5
0.75
1.25
1.5
1.75
2
Active and guided learning of enzyme function
Published on Oct 23, 20123126 Views
Manual annotation cannot keep up with enzyme sequence discovery. In this work, we modelled the use of active and guided learning to support enzyme function curation. We evaluated, on 5,750 E. coli
Related categories
Chapter list
Active and Guided Learning of Enzyme Function00:00
Predict enzyme function00:06
Enzyme Commission numbers00:24
Data sources - 101:00
Data sources - 202:00
Data schema - 102:04
InterPro sequence signatures03:07
InterPro signatures03:20
Data schema - 203:58
Algorithm04:16
SwissProt KEGG cross-evaluation05:33
Biocuration06:50
Dynamics of collaborative biocuration + machine learning07:21
Biocuration + machine learning07:31
Active learning08:18
Active learning: different from labelled & similar to unlabelled - 109:06
Active learning: different from labelled & similar to unlabelled - 209:25
Active learning: confidence - 109:38
Active learning: confidence - 209:56
Active learning disadvantages: nonparallel & poor on rare classes10:10
Distribution of Enzyme Commission numbers10:56
Guided learning tackle InterPro signature sets by frequency11:47
Current work: predict enzyme chemical mechanism16:08
MACiE: Mechanism, Annotation and Classification in Enzymes16:17
Mitchell's group16:32
Thank you17:15
Guided vs. Active Learning21:02