December, 2019 Newsletter 9
Content of this newsletter

  • BIGCHEM Farewell
  • Summary of ESRS fellows´outcomes
  • Publications
  • Presentations of BIGCHEM work
BIGCHEM Farewell

The BIGCHEM project finishes on December 2019, leaving great science, collaborations between industrial and academic partners and ten young scientists trained to lead the digital transformation of the drug discovery field. The BIGCHEM fellows’ work has already made a significant advancement in the field with the creation of novel in silico approaches that not only face presented challenges but also exploit the opportunities of Big Data to accelerate the development of new therapies.

The excellent scientific performance of BIGCHEM is well reflected in its prolific scientific publication outcome. Up to now, 41 peer-reviewed articles co-authored by BIGCHEM fellows and PIs, many in joint collaboration with academic and industrial partners, have been published in open access journals. Two students have already received their PhD degrees and others are working towards it. Furthermore, the BIGCHEM outcomes have been presented in numerous scientific conferences and events.

For this newsletter, the BIGCHEM ESR fellows have summarized their experience with BIGCHEM and their scientific achievements.
Summary of ESRS fellows´scientific outcomes
Raquel Rodríguez-Pérez
ESR4_Dipan_summary
"During the BigChem project, I have had the opportunity to investigate in academia and industry. In particular, I worked at the University of Bonn and Boehringer Ingelheim in Germany. My research has focused on one of the main goals of pharmaceutical research, which is the discovery of small molecules that bind to a biological target and modulate its activity. The experimental search for new bioactive compounds is time consuming and expensive. However, due to the increasing amounts of compound data that are currently available, there are high expectations associated with machine learning (ML) to prioritize experiments. My research has focused on the prediction of compound activity using ML models. I have predicted experimental screening results across a panel of biological targets using distinct strategies. Moreover, I have generated ML models to classify highly and weakly potent inhibitors for more than a hundred kinases. Accurate ML models can guide compound design and further experiments, but insights on predictions are required in practical applications. Hence, I have also introduced an intuitive approach to rationalize activity predictions from complex ML models. Interpretability enables to further validate the model based on domain knowledge and extract information from the chemical features driving the predictions. I have presented my research at different conferences, including the ACS National Meeting in Orlando, Florida (Spring 2019) and published six papers with my major findings."
Oliver Laufkötter
" My research has focused on diverse aspects of virtual database screening methods. My first project involved developing a novel method for representing compound descriptors/fingerprints. The aim of this project was to improve upon existing methods and to apply this method in machine learning studies focused around drug/lead discovery. Specifically, this method involved building a hybrid fingerprint which combined features of both chemical structure information and bioactivity profiles of compounds, termed BaSH fingerprint (Bio-activity Structural Hybrid). The bio-activity profiles stem from high throughput screening studies where millions of compounds are tested in hundreds of biochemical or cell-based assays. This bio-activity and structural hybrid descriptor was termed BaSH fingerprint. By hybridizing these two descriptor types a synergistic effect could be obtained which expressed an improvement in compound activity predictions in random forest machine learning models. My second research project involved benchmarking various similarity searching techniques and identifying which methods are best suited in virtual screening applications. A number of techniques were compared using a variety of parameter settings, including standard similarity searching, Turbo similarity searching, Tversky based similarity searching, and consensus bit scaling similarity searching. The outcome of the study showed that using methods with parameter settings which devalue features of reference compounds gave the best results, i.e. Tversky with a low alpha parameter or consensus bit scaling with low scaling factor. This led to the identification of a novel similarity searching technique termed negative consensus profile scaling. Two articles have been published based on these works, the publications detailing these studies are listed below. The figure shows the concept behind the first research project using the hybrid compound descriptor."
Dr. Arkadii Lin
Arkaii Lin carried out his PhD work at Boehringer Ingelheim Pharma Co. & KG (Germany) and University of Strasbourg (France). He developed new tools and methodologies to visualize, analyze and model Big Data in chemistry using the Generative Topographic Mapping (GTM) approach. Key applications of those approaches include the diversification of the in-house collection of Boehringer Ingelheim, multi-target virtual screening and the determination of optimal frame set size to describe the entire ChEMBL collection. His work has been published in 3 peer-review articles and presented in eight conferences. He successfully defended his PhD in September 2019 and afterwards, took a PostDoc position in the laboratory of Chemoinformatics in Strasbourg in collaboration with Janssen Global Service.
Dipan Ghosh
ESR4_Dipan_portrait
ESR4_Dipan_summary
Dipan PhD’s work at BIGCHEM has focused on the analysis of promiscuous compounds on chemical high throughput assays. Such potentially unwanted compounds should be identified and eliminated during the initial stages of screening. At HMGU, Dipan analyzed the assay interference problem due to the inhibition of luciferase, an enzyme widely used as a reporter. The work at the Lead Discovery Center centered on the identification on frequent hitters in G-protein coupled receptors (GPCR) assays. He has built robust machine learning models to identify promiscuous compounds in those assays and compared the performance with scaffold-based methods. All the models described were built using the OCHEM platform ( http://ochem.eu ) and are freely available to the public.
The Bigchem program gave me the opportunity to work in scientific research in both the academic world, in Prof. Dr. Jürgen Bajorath’s lab in Bonn, and the industrial world, in Dr. Ola Engkvist’s Molecular AI team in AstraZeneca. I learnt a lot about scientific research and the scientific community. On my PhD topic, which was the analysis of compounds promiscuity, I produced a publication focusing on the identification of compounds that interfere with HTS technologies. I also had the opportunity to work with the AstraZeneca team and Boehringer Ingelheim’s researchers on a review paper about deep-learning in pharma. This PhD was an exciting experience!”
Josep Arús-Pons
Josep_photo
" In these three years I have been working alongside Prof Jean-Louis Reymond in the University of Bern and with Dr. Ola Engkvist and Dr. Hongming Chen in AstraZeneca. The main goal of my work has been finding novel ways to explore the chemical space. While in Bern, I focused on creating enumerated molecular databases. Specifically the GDB4c database, which contained an enumeration of all ring systems with up to 4 rings. Further characterization of the database showed that 99% of the structures enumerated were novel, opening up possibilities in both synthetic and medicinal chemistry. On the other hand, in AstraZeneca I focused on exploring the chemical space using molecular deep generative models. These models are trained with a set of molecules (e.g. ChEMBL) and once trained are able to generate billions of molecules that are similar, but not equal, to those in the training set. Specifically, my research was understanding and improving models that use the SMILES syntax to represent molecules and to develop a benchmark to compare them. We were able to show that these models are able to easily generalize huge chemical spaces with very small samples. For instance, 83% of the GDB-13 database (containing 1 billion molecules) was learnt by a model trained with a set of only 0.1% of the database. All the results obtained during the PhD are a step in the direction of finding ways to explore unknown chemical space and thus to aid in the first steps of the drug discovery process. "
Dr. Xuejin Zhang (JJ)
JJ successfully finished her PhD work and defended her PhD thesis on 17 th September 2019 at ETH Hönggerberg.
Inspired by how human experts tackle the problem of reactant feasibility prediction for large combinatorial library design, Xuejin Zhang’s PhD’s work at BIGCHEM aimed to address this task by building a recommender. This recommender should be able to prioritize the reactants that bear compatible functional group, with high likelihood of being reacted for a reaction of interest.
"I investigated, developed, and validated methodological aspects of generating molecules with desirable properties. More specifically, I used different architectures of generative neural networks (NN) and researched how they can be fine-tuned to create novel molecules with certain specified properties. Currently, I authored / co-authored five publications. In the first studies, we introduced a method to tune a sequence-based generative NN for molecular de novo design that can learn to generate structures with specified desirable properties. We demonstrated how the model could execute a range of tasks such as generating analogs to a query structure and generating compounds predicted to be active against a biological target. In another work, we investigated the potential use of different NN autoencoders (AE) for de novo molecular design. An AE is a pair of NNs that maps a molecular structure into a continuous numerical representation, so-called latent space, and is able to reconstruct molecules from the latent space. The study showed that the latent space preserves the chemical similarity principle and thus can be used for the generation of analog structures. Furthermore, we studied what amount of chemical space can be sampled with generative models that are trained only with a very small sample of an enumerated database. We developed a method to assess the quality of the training process at different stages of the training and studied how much of the enumerated database can be reproduced by sampling the generative model at different stages of the training. Finally, we studied the ability of compounds to interact with multiple targets, so-called promiscuity. Various machine learning models were built to distinguish between promiscuous and nonpromiscuous compounds. The different machine learning models yielded overall successful predictions and provided evidence for the presence of structure−promiscuity relationships."
Computer Assisted Synthesis Planning (CASP) has gained considerable interest as of late. We have investigated CASP from two avenues, retrosynthetic route prediction, as an idea generator and reaction prediction as a form of validation, and likelihood estimator. Ultimately, the goal is to unite the two to provide an AI synthesis prediction tool capable of assessing the synthetic accessibility of a given target compound(s), whilst proposing routes with a literature precedent, and an assessment of the likelihood of failure. Such a tool has broad implications, in the form of time savings, efficient idea generation, and rapid assessment of synthesizability, thus reducing the overall failure rate. The retrosynthetic model which we have developed to date has been assessed on a series of internal virtual libraries and can predict the route to synthesized compounds with a ca. 70 % success rate on average. In the future, and as reaction prediction methods develop, the two approaches will be incorporated into one entity, providing a powerful tool for chemists, the automation lab, and de novo design groups to efficiently asses the synthetic accessibility of future potential small molecule therapies.
The PhD is currently ongoing and will continue for a period of six months at AstraZeneca, before returning to the University of Bern for the remaining years. It is hoped that as the tool develops, we are able asses how well predictions translate into the wet lab, and how they help researchers investigate synthesis in previously unexplored regions of chemical space.
MW_graphical_summary
At BIGCHEM, Michael has been working on the secure sharing of information using machine learning. His work began at HMGU with the Tetko group, focusing on molecular transformations and multitask learning, which resulted in two publications. During his subsequent stay at AstraZeneca, Michael worked on various contemporary deep learning algorithms under Dr. Hongming Chen and Dr. Ola Engkvist's supervision. His work used low-level molecular representations, and image data, to try to maximise information usage in activity and property prediction. This work has resulted in a conference paper and in another publication that is currently under peer review. Michael expects to submit and defend his thesis at the Technical University of Munich in early 2020.
Publications
  1. Rodríguez-Pérez R, Bajorath, J, Interpretation of Compound Activity Predictions from Complex Machine Learning Models Using Local Approximations and Shapley Values J. Med. Chem. 2019
  2. Engkvist O, Arús-Pous J, Bjerrum EJ, Chen H, Molecular de novo Design Through Deep Generative Models. Royal Society of Chemistry 2019, in press
  3. Arús-Pous J, Awale M, Probst D, Reymod JL, Exploring Chemical Space with Machine Learning. Chimia 2019, in press
  4. Arús-Pous J, Johansson S, Prykhodko O, Bjerrum EJ, Randomized SMILES strings improve the quality of molecular generative models. J Cheminform 2019, 11, 71
  5. Prykhodko, O; Johansson, SV; Kotsias, PC; Arús-Pous, J; Bjerrum, EJ; Engkvist, O; Hongming C, A De Novo Molecular Generation Method Using Latent Vector Based Generative Adversarial Network. ChemRxiv. 2019
  6. Blaschke T, Miljković F, Bajorath J, Prediction of Different Classes of Promiscuous and Nonpromiscuous Compounds Using Machine Learning and Nearest Neighbor Analysis. ACS Omega 2019, 4 (4), 6883-6890
  7. David L, Arús-Pous J, Karlsson J, Engkvist O, Bjerrum EJ, Kogej T, Kriegl J, Beck B, Chen H, Applications of deep-learning in exploiting large-scale and heterogeneous compound data in industrial pharmaceutical research. Front. Pharmacol. 2019, Nov 5
  8. Thakkar A, Selmi N, Reymond JL, Engkvist O, Bjerrum EJ, Ring Breaker’: Assessing Synthetic Accessibility of the Ring System Chemical Space. ChemRxiv
  9. Laufkötter O, Miyao T, Bajorath J, Large-Scale Comparison of Alternative Similarity Search Strategies with Varying Chemical Information Contents. ACS Omega 2019, Sep 5;4(12):15304-15311
  10. David L, Walsh J, Sturm N, Feierberg I, Nissink JWM, Chen H, Bajorath J, Engkvist O, Identification of Compounds That Interfere with High-Throughput Screening Assay Technologies. ChemMedChem 2019, Oct 17;14(20):1795-1802
  11. Thakkar A, Kogej T, Reymond JL, Engkvist O, Bjerrum EJ, Datasets and Their Influence on the Development of Computer Assisted Synthesis Planning Tools in the Pharmaceutical Domain. ChemRxiv
  12. Lin A, Beck B, Horvath D, Marcou G, Varnek A, Diversifying chemical libraries with generative topographic mapping. J Comput Aided Mol Des. 2019, Aug 12
  13. Thakkar A, Bjerrum EJ, Engkvist O, Reymond JL, Neural Network Guided Tree-Search Policies for Synthesis Planning. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  14. Arús-Pous J, Johansson S, Prykhodko O, Bjerrum EJ, Tyrchan C, Reymond JL, Chen H, Engkvist O, Improving Deep Generative Models with Randomized SMILES. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  15. Withnall M, Lindelöf E, Engkvist O, Chen H, Attention and Edge Memory Convolution for Bioactivity Prediction. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  16. Ghosh D, Tetko I, Klebl B, Nussbaumer P, Koch U, Analysis and Modelling of False Positives in GPCR Assays. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  17. Karpov P, Godin G, Tetko IV, A Transformer Model for Retrosynthesis. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  18. Tetko IV, Karpov P, Bruno E, Kimber TB, Godin G, Augmentation Is What You Need!. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  19. Lin A, Beck B, Horvath D, Varnek A, Diversify Libraries Using Generative Topographic Mapping. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  20. David L, Walsh J, Bajorath J, Engkvist O, Detection of Frequent-Hitters Across VariousHTS Technologies. In: Tetko I., Kůrková V., Karpov P., Theis F. (eds) Artificial Neural Networks and Machine Learning – ICANN 2019: Workshop and Special Sessions. ICANN 2019. Lecture Notes in Computer Science, vol 11731. Springer, Cham
  21. Laufkötter O, Sturm N, Bajorath J, Chen H, Engkvist O, Combining structural and bioactivity-based fingerprints improves prediction performance and scaffold hopping capability. J Cheminformatics, 2019 Aug 8;11(1):54
  22. Rodríguez-Pérez R, Bajorath J, Multitask Machine Learning for Classifying Highly and Weakly Potent Kinase Inhibitors. ACS Omega, 2019, 4 (2), 4367–4375
  23. Lin A, Horvath D, Marcou G, Beck B, Varnek A, Multi-task generative topographic mapping in virtual screening. J Comput Aided Mol Des, 2019, 33(3):331-343
  24. Arús-Pous J, Blaschke T, Ulander S, Reymond JL, Chen H, Engkvist O, Exploring the GDB-13 chemical space using deep generative models. J Cheminform, 2019, 11(1):20
  25. Sosnin S, Vashurina M, Withnall M, Karpov P, Fedorov M, Tetko IV, A Survey of Multi-Task Learning Methods in Chemoinformatics. Mol Inform. 2019, Apr;38(4):e1800108.
  26. Chen H, Engkvist O, Wang Y, Olivecrona M, Blaschke T, The rise of deep learning in drug discovery. Drug Discov Today, 2018, 23(6):1241-12502.
  27. Blaschke T, Olivecrona M, Engkvist O, Bajorath J, Chen H, Application of Generative Autoencoder in De Novo Molecular Design. Molecular informatics, 2018, 37 (1-2), 1700123
  28. Pinzi L, Caporuscio F., Rastelli G., Selection of protein conformations for structure-based polypharmacology studies. Drug Discov. Today, 2018, 23, 1889-1896
  29. Ghosh D, Koch U, Hadian K, Sattler M, Tetko IV, Luciferase Advisor: High-Accuracy Model To Flag False Positive Hits in Luciferase HTS Assays. J Chem Inf Model, 2018, 58 (5), 933-942
  30. Rodríguez-Pérez R, Miyao T, Jasial S, Vogt M, Bajorath J, Prediction of compound profiling matrices using machine learning. ACS Omega, 2018, 3 (6), 4713-4723
  31. Rodríguez-Pérez R, Bajorath J, Prediction of compound profiling matrices, part II: relative performance of multi-task deep learning and random forest classification on the basis of varying amounts of training data. ACS Omega, 2018, 3 (6), 12033-12040
  32. Lin A, Horvath D, Afonina V, Marcou G, Reymond JL, Varnek A, Mapping of the Available Chemical Space versus the Chemical Universe of Lead-Like Compounds. ChemMedChem, 2018, 13 (6), 540-554
  33. Withnall M; Chen H; Tetko I, Matched Molecular Pair Analysis on Large Melting Point Datasets: A Big Data Perspective. ChemMedChem, 2018, Mar 20;13(6):599-606.
  34. Rodríguez Pérez R, Vogt M, Bajorath J, Influence of Varying Training Set Composition and Size on Support Vector Machine-Based Prediction of Active Compounds. J. Chem. Inf. Model., 2017, Apr 24;57(4):710-7162.
  35. March-Vila E, Pinzi L, Sturm N, Tinivella A, Engkvist O, Chen H and Rastelli G, On the Integration of In Silico Drug Design Methods for Drug Repurposing. Front. Pharmacol., 2017, 8:298
  36. Olivecrona M, Blaschke T, Engkvist O and Chen H, Molecular De Novo Design through Deep Reinforcement Learning. J Cheminform, 2017, Sep 4;9(1):48
  37. Rodríguez-Pérez R, Vogt M, Bajorath, Support Vector Machine Classification and Regression Prioritize Different Structural Features for Binary Compound Activity and Potency Value Prediction. ACS Omega 2017, Oct 31;2(10):6371-6379
  38. Visini,R, Arús-Pous J, Awale M, Reymond JL, Virtual Exploration of the Ring Systems Chemical Universe. J. Chem. Inf. Model. 2017, Nov 27;57(11):2707-2718
  39. Awale M, Visini R, Probst D, Arus-Pous J, Reymond JL, Chemical Space: Big Data Challenge for Molecular Diversity. Chimia (Aarau) 2017, Oct 25;71(10):661-666
  40. Tetko, I. V.; Engkvist, O.; Koch, U.; Reymond, J. L.; Chen, H., BIGCHEM: Challenges and Opportunities for Big Data Analysis in Chemistry. Mol Inform 2016 , 35 (11-12) : 615-6212.
  41. Tetko, I.V.; Engkvist, O.; Chen, H. Does 'Big Data' exist in medicinal chemistry, and if so, how can it be harnessed?. Future Med Chem. 2016, 8(15):1801-1806
Presentations of BIGCHEM work
  • Biotechnological Forum, Moscow, Russia, February 2016
  • SLAS Compound Management Conference, Berlin, Germany, April 2016
  • 5th Chemoinformatics Strasbourg Summer School 2016, Strasbourg, France, July 2016
  • RICT 2016 - Interfacing Chemical Biology and Drug Discovery, Caen, France, July 2016
  • Second Kazan Summer School on Cheoinformatics, Kazan, Russia, July 2016
  • Seminar at Innopolis University, Innopolis, Russia, August 2016
  • American Chemical Soc (ACS) Fall conference 2016, Philladelphia, US, August 2016
  • European Federation of Medicinal Chemistry (EFMC) 24th International Symposium on Medicinal Chemistry, Manchester , UK, August 2016
  • XX Mendeleev congress, Yekaterinburg, Russia, September 2016
  • OpenTox meeting, Rheinfelden, Germany, October 2016
  • Society of Chemical Industry (SCI) Cheminformatics for Drug Design: Data, Models and Tools, Duxford, UK, October 2016
  • 11th German Conference on Chemoinformatics GCC 2016, Fulda, Germany, November 2016
  • Society of Chemical Industry (SCI) Highlights in Medicinal Chemistry II, London, UK, November 2016
  • German-Ukrainian Forum of Young Researchers, Kyiv-Lviv, Ukraine, December 2016
  • Drug Innovation in Academia, Heidelberg, Germany, December 2016
  • Shaping the Future: Big Data and Biomedicine and Frontier Technologies, Skoltech, Russia, April 2017
  • Seminar at the University of Bern, Bern, Germany, May 2017
  • Seminar at the University of Hannover, Hannover, Germany, June 2017
  • “RICT 2017 - 53rd International Conference on Medicinal Chemistry - Drug Discovery & Selection”, Toulouse, France, July 2017
  • “Big Data in Chemistry” STC 2017 conference, Basel, Switzerland, August 2017
  • Chemogenomics workshop in the University of Bonn, Bonn, Germany, August 2017
  • Swiss Chemical Society Fall Meeting; , Bern, Switzerland, August 2017
  • University of Bern, Bern, Switzerland, September 2017
  • 8es journées de la Société Française de Chémoinformatique, Orleans, France, October 2017
  • What can Big Data do for Chemistry?,, London, UK, October 2017
  • 13th German Conference on Chemoinformatics GCC 2017 conference, Mainz, Germany, November 2017
  • Helmholtz Structural Biology & SBDD, Düsseldorf, Germany, November 2017
  • Deutsch-Russischer Life Science Day Moscow, Moscow, Russia, February 2018
  • Artificial Intelligence Transforming Pharma R&D, Boston, US, February 2018
  • 32nd Molecular Modelling Workshop, Erlangen, Germany, March 2018
  • Seminar at Skoltech, Moscow, Russia, March 2018
  • SLAS Compound Management Conference, Berlin, Germany, March 2018
  • Third International School-Seminar “From Empirical to Predictive Chemistry”, Kazan, Russia, April 2018
  • International Conference on Computational Chemistry and Toxicology, Taichung, Taiwan, May 2018
  • 6th Strasbourg Summer School in Chemoinformatics, Strasbourg, France, June 2018
  • Summer School on Machine Learning in Drug Design MLDD 2018, Leuven, Belgium, August 2018
  • Chemogenomics workshop in the University of Bonn, Bonn, Germany, September 2018
  • Chemoinformatics Workshop, Lausanne, Switzerland, September 2018
  • Swiss Chemical Society Summer Conference, Laussanne, Switzerland, September 2018
  • Artificial Intelligence in Chemical Research, Stein, Switzerland, October 2018
  • Institute of Bioengineering of Catalonia IBEC Symposium, Barcelona, Spain, October 2018
  • Invited lecture at Beijing University of Chemical Technology, Beijing, China, November 2018
  • 14th German conference on chemoinformatics GCC 2018, Mainz, Germany, November 2018.
  • OpenRiskNetMmeeting, Brussels, Belgium, December 2018
  • SLAS Compound Management Conference, Berlin, Germany, March 2019
  • ACS National Meeting, Orlando, US, April 2019
  • Chemoinformatics Workshop B-IT, Bonn, Germany, April 2019
  • Merck Science Dialogue, Moscow, Russia, April 2019
  • Seminar at the Institute of Computational Biology, Neuherberg, Germany, April 2019
  • First Israel-French Workshop, Ramat Gan, Israel, May 2019
  • Life science informatics programm at university of Bonn, Bonn, Germany, May 2019
  • PhysChem Forum PCF18, Frankfurt, Germany, May 2019
  • 47th conference of the International Union of Pure and Applied Chemistry, Paris, France, July 2019
  • RICT2019 conference, Nantes, France, July 2019
  • “ICANN 2019” International Conference on Artificial Neural Networks, Munich, Germany, September 2019
  • EFMC Synthetic and Medicinal Chemistry, Athens, Greece, September 2019
  • Biopharmaceuticals R&D Science Symposium, Gothenburg, Sweden, October 2019
  • Seminar at the University of Mainz, Mainz, Germany, October 2019
  • "Beyond the Rule of 5", Nottingham, UK, November 2019
  • 9th conference of the French Society of Cheminformatics, Paris, France, November 2019
  • Berlin Science Week, Berlin, Germany, November 2019
  • ICAIH 2019 - First Industrial Conference on Artificial Intelligence and Health, Milano, Italy, November 2019
  • German Cheminformatics Conference, Mainz, Germany, November 2019
  • 5th Workshop of Structural Biology, Heidelberg, Germany, December 2019
  • AI Powered Drug Discovery and Manufacturing , Cambridge, MA, US, February 2020
BIGCHEM Newsletters
You can access previous newsletters at the BIGCHEM web site.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 676434. 
Disclaimer: the newsletter reflects only the authors’ view and neither the European Commission nor the Research Executive Agency are responsible for any use that may be made of the information it contains.