Computational Chemistry Days 2018 - Computational Chemistry Days 2018
Computational Chemistry Days 2018 - connecting chemists and materials scientists in Finland and beyond
The Computational Chemistry Days 2018 were organised by the University of Helsinki on 28th–29th May. The Days have a long, successful history of bringing scientists together to initiate collaboration and share inspiring ideas. I'd like to express my warmest thanks to the organizers for creating a relaxed, open and friendly atmosphere while practical arrangements worked smoothly. Next year, the Days will take place in Kuopio.
The days were attended by over 100 people, the majority from Finnish universities, with ca 20% from elsewhere in Europe. The event was a great way to discover what's going on in other universities and make connections for future collaborations, especially so for the student participants, who made up half of the audience.
Emphasis on method development including machine learning
This year the emphasis was on computational method development, including machine-learning techniques. Since computational methods have been used in chemistry for decades the variety is large and therefore it was both useful and interesting to see the advances on multiple fronts. I also believe exposure to latest methods outside one's own bubble (or de facto standard tools) will be useful to make big steps forward. In fact, if a new bleeding-edge-method is found, likely being developed in a group somewhere in Europe, the researcher might want to make use of HPC-Europa3 (see below) to visit the group to facilitate adopting the method in one's own research. The city of Helsinki had arranged a welcome reception in the Bock house hosted by Deputy Mayor for Education Pia Pakarinen, and Senior Adviser to the Mayor Victor Andersson.
Some highlights of talks
In his plenary talk, Prof. Roland Lindh from Uppsala outlined how quantum chemistry has evolved from its roots at the beginning of the last century, and the changes that artificial intelligence and machine learning will bring in the current century. Predictive power of computational methods will grow significantly enabling completely new approaches. One is the inverse design problem, where instead of simulating the properties of a given molecule, one simulates what kind of molecule would possess a given property. Prof. Lindh also cited a number of other challenges, one of which is the Chemical Turing Test; if and when the results of a simulation become indistinguishable from the results of an actual physical experiment, choosing whether to do theory or experiment boils down to factors like cost and environmental sustainability.
Prof. Patrick Rinke from Aalto University presented the benefits of sharing your computational results in the NOMAD CoE (Novel Materials Discovery Laboratory) in his talk. The NOMAD services automatically parse and add metadata to the raw output of your calculations and make them discoverable by others - or by you: how much of the data you or your post docs generated two years ago can you still find? The NOMAD tools allow easy browsing and filtering of the data, as well as more advanced machine learning analysis. Prof. Rinke also presented a new genealogy or tree of life of materials that will be used to categorize uploaded content in NOMAD.
One of my personal favourites was Prof. Esa Räsänen's overview of signal processing using the DFA (Detrended Fluctuation Analysis) and its applications in heart monitoring but also on explaining why Jeff Porcaro's groove is so famous. The connection to chemistry is a bit elliptic, but I'm sure there are always signals to be analysed and DFA seems like a robust and informative alternative.
It's important to meet researchers in person
Personally I find the face to face discussion with researchers always very useful but also interesting. During the event, the discussions resulted immediately in a few improvements in CSC's services and software portfolio. Thank you for the useful discussions! In retrospect, this is not a big surprise, since the participants came from an extremely relevant customer segment for CSC. Most (all except one?) talks given by the researchers affiliated at Finnish universities had used CSC's resources (and appropriately acknowledged it, thanks!). Correspondingly, CSC seems to be an important infrastructure for the computational chemistry community in Finland. CSC recently conducted a customer survey, whose results have been analysed and actions are being implemented. Take a look at a blog entry summarising the feedback (link below).
In addition to conversing with the participants, we also strived to improve the researchers' awareness of the HPC-Europa3 programme which funds international research visits involving computational research. Based on the statistics of accepted visits, chemists and materials scientists have been the most active fields of science grabbing this opportunity. Despite this, HPCE3 could support many more visits, both to Finland but also from Finland to Europe. If you want to start new or strengthen existing collaboration, why not let the European Union cover the costs? If you have any questions, don't hesitate to contact us at email@example.com.
Poster session gate crashed by virtual reality
The poster session gathered almost 30 posters competing for two prizes: one voted by all participants and another selected by an expert jury. As one of the organizers Dr. Mikael Johansson aptly noted in the prize ceremony, the field is evidently progressing as one of the poster prizes was awarded not to a traditional 2D printed sheet at all, but to an interactive hands-on virtual reality experience "VRChem: A molecular modeling software for Virtual Reality" by Otso Pietikäinen from Aalto University. With VRChem the researcher can build and manipulate chemical structures using a HTC Vive set. The other poster award was given to Dr. Piotr Batys et al. from Aalto University for their work on where the unique properties of spider silk come from (more details on these can be found in the Abstract book available on the event website, see below).
Top image: Prof. Roland Lindh outlining the development of computational chemistry and anticipating great advances during the next years. Photo: Nino Runeberg, CSC.
Published originally 1.6.2018.
Author works in customer support including chemistry and international projects.