2021 Fiscal Year Research-status Report
Use of contact prediction-based restraints for protein structure determination from sparse NMR data
| Project/Area Number |
20K06508
|
|---|
| Research Institution | Tokyo Metropolitan University |
|---|
Principal Investigator |
PETER GUENTERT 東京都立大学, 理学研究科, 客員教授 (20392110)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
池谷 鉄兵 東京都立大学, 理学研究科, 准教授 (30457840)
伊藤 隆 東京都立大学, 理学研究科, 教授 (80261147)
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | machine learning / NMR / protein structure / automated assignment |
|---|
| Outline of Annual Research Achievements |
In the past fiscal year, we worked on improving the machine learning-based fully automated NMR spectra analysis method. Using only NMR spectra and the protein sequence as input, our machine learning-based method delivers signal positions, resonance assignments, and structures strictly without any human intervention. Tested on a 100-protein benchmark, it demonstrated its ability to solve structures with 1.44 Angstrom median RMSD to the PDB reference and to identify 91.36% correct NMR resonance assignments. The method effectively reduces the effort for a protein assignment or structure determination by NMR essentially to the preparation of the sample and the NMR measurements.
In this project, this method will be combined with predictions that are made by the machine learning-basedAlphaFold2 software purely on the basis of the protein sequence in order to establish a hybrid method for NMR assignment and structure determination that can deliver experimentally derived assignments and structures using significantly smaller, and thus much faster measured sets of input NMR spectra.
Crucially for the present project, AlphaFold2 has now become freely available, and can be employed in our research. AlphaFold2 predicts structures of proteins based on their sequence with much better accuracy than any previous approach. We have installed AlphaFold2 on our local computer systems, applied it to all proteins for which we have NMR spectral data suitable for automated analysis, and initiated work on introducing the AlphaFold2 predictions into our fully automated NMR pipeline.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
The free availability of the AlphaFold2 software for sequence-based protein structure prediction with unprecedented reliability, enabled us to make direct use of this crucial method in our project. In the previous fiscal year, AlphaFold2 had been published but it remained unclear whether and when it would become possible to employ it in our project. This uncertainty has been resolved completely.
Research was still negatively affected by travel restrictions due to the COVID-19 pandemic, which have barred the main applicant from traveling to Japan.
|
| Strategy for Future Research Activity |
We plan to further investigate several different methods to introduce AlphaFold2 predictions into our fully automated NMR pipeline. Data from AlphaFold2 can be incorporated in the input for (i) automated chemical shift assignment, (ii) automated NOESY-based distance restraint determination, and (iii) structure calculation. For (i), the AlphaFold2-predicted structure serves to extend the list of expected NOESY peaks by those that correspond to short distances that are long-range with respect to the protein sequence. For (ii), the iterative procedure of combined NOESY distance restraint assignment and structure calculation is started from the AlphaFold2-predicted structure rather than from no structure at all. For (iii), the coordinates of the atoms are restrained by weak positional restraints towards the structure predicted by AlphaFold2. These positional restraints are applied with such low weights that the experimental NMR data can easily override them, while they help to keep the structure in place in regions where insufficient NMR data is available.
|
| Causes of Carryover |
今年度は，コロナ禍による影響のため，当初予定していた出張費等でほとんど支出がなかった．次年度は，コロナ禍の問題も解消され，外部研究者との共同研究やディスカッションも積極的に行う予定である．また今年度までに開発してきたソフトウェアを実際のデータに応用するための計算機を購入予定である．
|
-
[Journal Article] Atomic-resolution chemical characterization of (2x)72 kDa tryptophan synthase via 4D and 5D 1H-detected solid-state NMR2022
Author(s)
Klein, A., Rovo, P., Sakhrani, V. V., Wang, Y., Holmes, J. B., Liu, V., Skowronek, P., Kukuk, L., Vasa, S. K., Guentert, P., Mueller, L. J., Linser, R.
-
Journal Title
Proceedings of the National Academy of Sciences of the United States of America
Volume: 119
Pages: e2114690119
DOI
Peer Reviewed / Open Access / Int'l Joint Research
-
-
-
[Journal Article] Paramagnetic solid-state NMR to localize the metal-ion cofactor in an oligomeric DnaB helicase2021
Author(s)
Zehnder, J., Cadalbert, R., Terradot, L., Guentert, P., Boeckmann, A., Meier, B. H., Wiegand, T.
-
Journal Title
Chemistry Europe
Volume: 27
Pages: 7745-7755
DOI
Peer Reviewed / Open Access / Int'l Joint Research
-
[Journal Article] 13C and 15N resonance assignment of the YTH domain of YTHDC22021
Author(s)
He., F., Endo, R., Kuwasako, K., Takahashi, M., Tsuda, K., Nagata, T., Watanabe, S., Tanaka, A., Kobayashi, N., Kigawa, T., Guentert, P., Shirouzu, M., Yokoyama, S. & Muto, Y.
-
Journal Title
Biomolecular NMR Assignments
Volume: 15
Pages: 1-7
DOI
Peer Reviewed / Int'l Joint Research
-
-
-
-
