Empirical models
From Proteopedia
(Difference between revisions)
| Line 1: | Line 1: | ||
| - | ''Empirical models'' of molecules are those obtained by experimental observation of the molecules themselves. They are usually the most reliable structures of macromolecules. The [[Protein Data Bank]] (PDB) is the international archive of empirically determined models. The most common experimental methods are evident from the numbers of each kind of experimental result in the PDB (data from February 18, 2021, when the total number of entries in the PDB was 174,826): | + | ''Empirical models'' of molecules are those obtained by experimental observation of the molecules themselves. They are usually the most reliable structures of macromolecules. The [[Protein Data Bank]] (PDB) is the international archive of empirically determined models. The most common experimental methods are evident from the numbers of each kind of experimental result in the PDB (data from '''February 18, 2021''', when the total number of entries in the PDB was 174,826): |
* [[X-ray crystallography]]: 154,329 (88%) | * [[X-ray crystallography]]: 154,329 (88%) | ||
* [[NMR|Solution Nuclear Magnetic Resonance]]: 13,210 (7.6%) | * [[NMR|Solution Nuclear Magnetic Resonance]]: 13,210 (7.6%) | ||
Revision as of 21:33, 21 January 2022
Empirical models of molecules are those obtained by experimental observation of the molecules themselves. They are usually the most reliable structures of macromolecules. The Protein Data Bank (PDB) is the international archive of empirically determined models. The most common experimental methods are evident from the numbers of each kind of experimental result in the PDB (data from February 18, 2021, when the total number of entries in the PDB was 174,826):
- X-ray crystallography: 154,329 (88%)
- Solution Nuclear Magnetic Resonance: 13,210 (7.6%)
- Electron microscopy: 6,832 (3.9%)
- Other methods: 459 (0.3%)
Empirical models are distinguished from Theoretical models. In the past, theoretial models have been unreliable. However, in 2020 AlphaFold made a surprising breakthrough, and has been able to predict many structures very reliably.
