This is a modal window.
Beginning of dialog window. Escape will cancel and close the window.
End of dialog window.
Lecture Computational Proteomics and Metabolomics, 20. LessonKohlbacher, Oliver , 13.01.2015Table of Contents: |
|
00:00:00 | Protein Ambiguity Groups, Reporting |
00:02:39 | Inference through Quantification |
00:08:23 | Inferred Hits, Significance |
00:10:18 | Single Hit Wonders |
00:12:24 | Target-Decoy Databases (Recapitulation) |
00:13:08 | FDR Calculation (Recapitulation) |
00:13:32 | Single Hit Wonders |
00:16:51 | Protein FDRs |
00:19:12 | ProteinProphet |
00:21:06 | PeptideProphet, Peptide Probability Estimates (PPE) |
00:22:24 | Protein Probability Estimates |
00:27:36 | ProteinProphet (Example) |
00:29:50 | Sibling Peptides |
00:31:51 | NSP Distributions |
00:32:56 | NSP Correction, Influence |
00:33:30 | Protein Ambiguity |
00:35:43 | Protein Ambiguity Group |
00:36:45 | Protein FDR Calculation |
00:39:25 | MAYU |
00:42:26 | MAYU vs. ProteinProphet |
00:44:03 | MAYU |
00:46:17 | Benchmarking Inference Engines |
00:47:59 | Lecture Computational Proteomics and Metabolomics, Learning Units 9A, 9B and 9C: Materials and References |
Autoplay | |
Quality | 480p |
Playbackrate | 1.0 |
Embedded timms Player |
||||
Mode | Size | Settings | ||
iframe | ||||
Sorry, video window to small to embed... |
|
|
||||