Read on Twitter
Here it is ! New article on calculation of #NMR parameters of Starchy materials
https://pubs.acs.org/articlesonrequest/AOR-JVRVB5WYVQSPCZBSSQJD
A short thread of what's inside : 1/n
https://pubs.acs.org/articlesonrequest/AOR-JVRVB5WYVQSPCZBSSQJD
A short thread of what's inside : 1/n
During my Ph.D I was insterested in structuration of starch in presence or absence of small ligands. Basically starch may form 2 helical structures in aqueous environment : one is a double helix ( A or B polymorph) the other one is a single helix encapsulating a guest molecule.
In this paper we were interested in the B-polymorph. The difference between A and B resides in the cristalline packing. The aim of all calculations that I've performed was the reproduction of the NMR spectra of this double helix. The point is : this is an annoying NMR spectrum.
Indeed, it's been shown in 2006, if starch is composed only of one kind of sugar residues (glucopyranosyl ones), in the case of the B polymorph, these residues show 2 different NMR signals
https://pubs.acs.org/doi/abs/10.1021/bm060330x
https://pubs.acs.org/doi/abs/10.1021/bm060330x
We somehow took this as a challenge and we decided to try to reproduce this feature 
We then genuinely started what everybody tries in such case : Geometry optimisation -> Chemical shielding calculation. We first used molecular methods (DFT, GIAO).
We then genuinely started what everybody tries in such case : Geometry optimisation -> Chemical shielding calculation. We first used molecular methods (DFT, GIAO).
At this point I should mention that our initial structure comes from the @CermavDirection and early work of @AnneImberty . Thank you for letting everyone access to the structures you published !
http://polysac3db.cermav.cnrs.fr/db-connect.php?number=11 https://onlinelibrary.wiley.com/doi/abs/10.1002/bip.360270803
http://polysac3db.cermav.cnrs.fr/db-connect.php?number=11 https://onlinelibrary.wiley.com/doi/abs/10.1002/bip.360270803
Geometry optimisations led us to several structures with differents hydrogen bond networks(HBN). And each of these HBN displayed different chemical shifts. We performed molecular dynamics simulations and could show that these HBNs may swap from one to another.
The swap between these structures may happen rapidly enough to be in the situation of fast exchange, which is important to consider in such case, but this is not the point here. The main point was : We couldnt differentiate 2 environment and so 2 NMR signals in our calculations
Here we then contacted one of our colleague I. Gerber from #LPCNO Toulouse, which is an expert of solid-state physics. And we decided to add periodical boundaries to the optimisation and to the NMR calculation ( GIPAW method, many thx Pickard & Mauri, your method rocks !)
And here are the results ( which I find personnally beautiful
): the residues are colored accordingly to their chemical shift. The correlation with their placing in the cristal is becoming obvious.
): the residues are colored accordingly to their chemical shift. The correlation with their placing in the cristal is becoming obvious.
The influence of the cristalline packing has been postulated earlier by experimentalists, and we could show that in this case it is very important to take it into account in our calculations. Here, an example of interaction you miss if you don't include periodicity :
The NMR results we obtained are not perfect, since we miss out some effects (water molecules especially) but the main point of this paper is to sum up and prove that there are some ingredients you have to take into account if you want to calculate NMR spectra of these compounds.
Many thx to @fra_jo31 and V. RĂ©at (not on twitter ?) for leading me through my Ph.D and a tremendous mentoring work, and to I. Gerber for his help.
Poke @ElisaTelisa @SebVidalChem @DelbiancoM sry for pure shameless self-promotion but it may interest all of you.
Poke @ElisaTelisa @SebVidalChem @DelbiancoM sry for pure shameless self-promotion but it may interest all of you.
( @SebVidalChem please don't vidalize me, I hope there won't be any mistakes in representations of sugars, but I can't be sure 
)
)
