Thursday 20th April 2023 / 9:00 AM – 17:00 PM / Wellcome Genome Campus, Hinxton, Cambridgeshire
The Spring UKQSAR & Cheminformatics Group Meeting will be held at the Wellcome Genome Campus, Hinxton, Cambridge, on Thursday 20th April 2023. The meeting is organised jointly by EMBL-EBI and Sosei Heptares, and as always is free. The theme this time is “Learning from data” and for the occasion three of the most relevant databases for our fields will be introduced (PubChem, ChEMBL and the Cambridge Structural Database). The afternoon sessions will be focused on protein structure-based techniques (including use of AlphaFold, and ML for virtual screening) and reaction informatics (including applications of Enamine REAL, and machine learning).
LATEST UPDATE
**REGISTRATION TO ATTEND IN PERSON IS NOW CLOSED**
- Join the waiting list to be notified if a place becomes available, or register virtually to join the meeting online.
- Presentation abstract list has been updated
- Posters: We are delighted to announce that for this meeting the poster winner will receive a free waiver to publish an article in Journal of Cheminformatics.
- Transport: We have arranged a return shuttle bus to transport delegates traveling from Whittlesford Parkway Train Station to EMBL-EBI on Thursday 20 April 2023. More details when you register.
The Spring UKQSAR & Cheminformatics Group Meeting will be held at the Wellcome Genome Campus, Hinxton, Cambridge, on Thursday 20th April 2023. The meeting is organised jointly by EMBL-EBI and Sosei Heptares, and as always is free. The theme this time is “Learning from data” and for the occasion three of the most relevant databases for our fields will be introduced (PubChem, ChEMBL and the Cambridge Structural Database). The afternoon sessions will be focused on protein structure-based techniques (including use of AlphaFold, and ML for virtual screening) and reaction informatics (including applications of Enamine REAL, and machine learning).
REGISTRATION
Please register using the following link: https://www.eventsforce.net/embl/93/register
Only registered participants will be allowed to attend.
Registration closes at 23.30 BST on the 10th April.
LOCATION AND TRAVEL
The EMBL-EBI building is located on the Wellcome Genome Campus. More details on the campus are available here.
EMBL-EBI, Wellcome Genome Campus, Hinxton, Cambridgeshire, CB10 1SD, UK
Please find information on traveling to EMBL-EBI on our website.
AGENDA
| 09:00 | Open registration, coffee/tea | |
| 10:15 | Welcome and opening remarks | Steve Maginn (UKQSAR), Andrew R Leach (EMBL-EBI) |
| Session 1: Data for advanced research – Chair: Nicolas Bosc (EMBL-EBI) | ||
| 10:30 | PubChem: a wealth of information for your research | Evan Bolton (PubChem) |
| 11:00 | EMBL-EBI’s Chemical Biology Resources for Drug Discovery – what does the future hold? | Barbara Zdrazil (EMBL-EBI) |
| 11:30 | Learning from the Cambridge Structural Database | Isaac Sugden (CCDC) |
| 12:00 | Lunch and poster session | |
| Session 2: Learning from data – Protein structure-based techniques – Chair: Chris de Graaf (Sosei Heptares) | ||
| 13:30 | Applications of AlphaFold and Generative Molecular Design for GPCR SBDD | Brian Bender (Sosei Heptares) |
| 14:00 | Expanding the target coverage and accuracy of virtual screening in the age of machine learning | Marton Vass (Schrodinger) |
| 14:30 | Break and poster session | |
| Session 3: Learning from data – Reaction informatics – Chair: Noel O’Boyle (Sosei Heptares) | ||
| 15:00 | Making virtual REAL: creation and application | Yurii Moroz (ChemSpace) |
| 15:30 | Optimizing catalytic reactions using machine learning | Lucy van Dijk (Vertex) |
| 16:00 | Machine Learning around the Design-Make-Test Cycle | Marwin Segler (Microsoft) |
| 16:30 | Concluding remarks, poster prize | Chris de Graaf/Noel O’Boyle |
ABSTRACTS & PRE-READING MATERIAL
- Evan Bolton, Head Chemistry Program (National Library of Medicine)
PubChem: a wealth of information for your research”
PubChem is a massive resource. An effective strategy to access what you need requires a little knowledge and imagination. This talk will give you a broad overview of major ways to access information from PubChem whether you are a beginner or an expert or whether you are a technophobe or a technophile. In addition, this talk will give you some illustrative examples to help you get a sense of how PubChem can help meet your research needs.
Suggested reading
https://doi.org/10.1016/j.jmb.2022.167514
https://doi.org/10.1002/cpz1.217
https://doi.org/10.1093/nar/gkaa971
- Barbara Zdrazil, ChEMBL Coordinator, EMBL-EBI
EMBL-EBI’s Chemical Biology Resources for Drug Discovery – what does the future hold?
The Chemical Biology Services’ Team at the European Bioinformatics Institute manages the ChEMBL database of quantitative small-molecule bioactivity data used in drug discovery; SureChEMBL, for patent information; UniChem, a convenient way to cross-reference multiple databases using chemical structure; and ChEBI, a database of small molecules relevant to biology. All our resources are open and follow FAIR principles. Recognising their global importance, ChEMBL and ChEBI have recently been designated as Global Core Biodata Resources (https://globalbiodata.org/scientific-activities/global-core-biodata-resources/).
Over time these resources have grown in complexity and size, and they have been used by the drug discovery and related communities to answer practical science questions. Example applications include the identification of tool compounds for potential therapeutic targets, novelty evaluation, target profiling for bioactive compounds, and many more. ChEMBL is also very widely used in data science and in the development, validation, and application of AI/ML methods.
In this talk, we will describe the evolution of our resources, provide some practical examples of their usage, and outline some of the new functionalities and plans for the future.
- Isaac Sugden, Computational and Data Science Team Leader (CCDC)
Learning from the Cambridge Structural Database
The world of structural chemistry reached a tremendous milestone in 2019, with the publication of over one million organic and metal organic crystal structures. Each one of these, now, 1.2 million plus structures is shared through the Cambridge Structural Database (CSD) and is something the entire scientific community can be proud of. Today the CSD and the accompanying software are used by scientists worldwide working in over 70 countries, and the knowledge derived from the structural data contained in the CSD has underpinned fundamental chemical discoveries and played a key role in designing new materials from drugs to semiconductors and beyond.
This talk will look at some of the research that the CSD has enabled and highlight the knowledge and insights that have been derived from this collection of data. It will explore how the database has evolved over the years and how researchers have used AI and machine learning techniques on this valuable source of data. 2 particular case studies will be presented: focussing on how structure activity relationships within the CSD were used to predict novel semiconductors, and successfully rank the conformers that may form cocrystals with a given API.
- Márton Vass, Principal Applications Scientist (Schrodinger)
Expanding the target coverage and accuracy of virtual screening in the age of machine learning
Machine learning technologies are now influencing all stages of the drug discovery process. Since the advent of AlphaFold2 it has been an open question whether this or derived methods could provide starting points for structure-based drug discovery techniques. We have investigated the usability of AlphaFold2 models in virtual screening experiments using the unrefined models or their binding sites refined with Schrödinger’s IFD-MD tool. We show that unrefined structures, on average, provide enrichments on par with apo crystal structures, whereas refined structures can provide enrichments close to their holo counterparts. A combination of IFD-MD and FEP technologies can also be used on homology models and AlphaFold2 structures to uncover compound binding modes in novel targets and for structure-based hit expansion and lead optimization.
Suggested reading
https://chemrxiv.org/engage/chemrxiv/article-details/62b41f0c0bbbc117477285a4
https://pubs.acs.org/doi/abs/10.1021/acs.jctc.1c00136
https://pubs.acs.org/doi/10.1021/acs.jctc.2c00371
- Yurii Moroz, CEO (ChemSpace)
Making virtual REAL: creation and application.
Accessible chemical spaces play a role in early drug discovery. The main criteria for molecules from the space include novelty, deliverability, and accessibility. Enamine REAL Space has been a leader among accessible chemical spaces delivering dozens of successful stories, both published and proprietary. A few approaches have been published describing ways to explore the REAL Space. The current presentation will describe the workflow for the REAL Space creation and expansion as well as shows use cases of its application to advance drug discovery projects.
Suggested reading
https://doi.org/10.1021/acs.jmedchem.2c00813
https://doi.org/10.1038/s41586-021-04220-9
https://doi.org/10.1016/j.isci.2020.101681
FOR POSTER PRESENTERS
As part of the registration process, you will be offered to submit an abstract. This will be reviewed by the organising committee, who will then confirm whether you will be able to present a poster. If you would like to present a poster, please register as soon as you can to allow us time to review your abstract and you time to create your poster before the day.
Key dates and information:
- Abstracts can only be submitted to the organising committee at the time of registration
- Abstracts must NOT EXCEED 300 CHARACTERS, and must be uploaded as .pdf files
- You will be notified by 5pm on 12TH APRIL if you are successful
- All posters need to be PRE-PRINTED as A0 (33.1 inches x 46.8 inches / 841 mm x 1189 mm) or A1 size (23.4 x 33.1 inches / 594 x 841 mm)
- Posters must be brought along ready for 9AM BST on the day of the meeting to allow enough set-up time
- The event organisers will attach all posters to the boards on the day
HOTELS
We recommend the following hotels. In distance order from the venue, starting with the closest first.
Hinxton Hall Conference Centre (located on campus)
Please note that Hinxton Hall has limited availability for this date so delegates are advised to book as soon as possible.
Red Lion Hinxton (5 minute walk from campus)
HOLIDAY INN EXPRESS CAMBRIDGE-DUXFORD M11, JCT.10 (a short taxi ride away from campus)
For more information on accommodation please see our website here.