• Guntram Bauer

• Minoru Kanehisa

Since the Human Genome Project (HGP) in the 1990s, continuous developments of high-throughput experimental technologies and generation of large-scale datasets have transformed biology into data science. The KEGG database resource (https://www.kegg.jp) was initiated in 1995 under the Japanese HGP and we share our perspectives on its past, present and future developments. KEGG is not a simple data repository. KEGG is a computerized model of biological information systems in the cell, the organism and the ecosystem, which enables uncovering hidden features in genome sequences and other biological data. In comparison to AI/ML models that are computationally generated from big data, the KEGG model is manually created from selected publications based on human intelligence. This aspect never changed over the past 28 years, but there were two major changes. One is the change from a publicly funded database to a self-sustaining database, in which the licensing revenue received from commercial users is fully reinvested to further develop KEGG and to continue to make it freely available to academic users. The other change involves more focus on social values. The health information category of KEGG was introduced in 2010 around the time when public funding started to decline. It integrates drug labels and other regulatory data with scientific knowledge, and KEGG is now a popular web resource for drug information in the Japanese society. The essence of the KEGG model is its molecular network representation, which we hope will be expanded enabling the analysis of molecular reactions and interactions at the biosphere level.

• Johanna McEntyre

The mission of the EMBL-EBI data services is to deliver global data resources that meet the needs of cutting edge life sciences and medical research, and to do so in the most cost-effective, integrated, and sustainable way possible. Molecular and related data are provided to the world from just over 40 data resources, covering genomes to chemicals to curated cellular pathways and the research literature. These resources are used everyday use in the design and interpretation of experiments, through to supporting sophisticated new analyses schemes that push forward the boundaries of understanding, to routine use in clinical genomics workflows worldwide and finally in enabling transformative computational approaches, such as DeepMind’s AlphaFold. In this presentation I will describe how this portfolio of resources is managed, highlighting some key recent developments, challenges and future opportunities.

• Amber Scholz, Leibniz Institute DSMZ, Germany

The UN Convention on Biological Diversity establishes sovereign rights of nations over the biological diversity in their borders. From pathogens to palm trees, countries can regulate the terms of access and benefit-sharing for their genetic resources under the Nagoya Protocol. But what happens when scientists share the data on these same organisms in open access databases and infrastructures? Who should benefit from these data and how can benefits flow back to the original country of origin? Recent international policy decisions under the Kunming-Montreal Global Biodiversity Framework and on-going discussions under the Food and Agriculture Organization, World Health Organization and UN Convention on the Law of the Seas will fundamentally change how biological data ownership is handled. The decisions will likely impact biological database management principles, shift data stewardship practices, and revise the ethical standards of the bioinformatics community into the next decade and beyond. Come learn more about policy decisions that will impact your understanding of data ownership, ethics, and science policy.

• Josefina Campos

• Halima Bensmail, Qatar computing research institute, Hamad Bin Khalifa University, Qatar

Monkeypox virus (MPXV) was confirmed in May 2022 and designated a global health emergency by WHO in
July 2022. MPX virions are big, enclosed, brick-shaped, and contain a linear, double-stranded DNA genome
as well as enzymes. MPXV particles bind to the host cell membrane via a variety of viral-host protein
interactions. As a result, the wrapped structure is a potential therapeutic target. DeepRepurpose, an artificial
intelligence-based compound-viral proteins interaction framework, was used via a transfer learning
setting to prioritize a set of FDA approved and investigational drugs which can potentially inhibit MPXV
viral proteins. To filter and narrow down the lead compounds from curated collections of pharmaceutical
compounds, we used a rigorous computational framework that included homology modeling, molecular
docking, dynamic simulations, binding free energy calculations, and binding pose metadynamics. We
identified Elvitegravir as a potential inhibitor of MPXV virus using our comprehensive pipeline.

• Nicola Mulder, University of Cape Town, South Africa

New and improved technologies are resulting in biomedical data being generated at unprecedented scales. In order to convert the data into knowledge for application to health priorities, it is necessary to integrate different data types and embrace data science techniques for their analysis and interpretation. This increased complexity in biomedical research brings with it many challenges, particularly for low resourced settings, such as those in low and middle income countries. Challenges include moving, storing, securing and processing large datasets and doing so in an ethical, responsible way such that those with authorized access to data have access to the required data analysis tools and computing resources. The Harnessing Data Science for Health Discovery and Innovation in Africa (DS-I Africa) Initiative aims to “leverage data science technologies to transform biomedical and behavioral research and develop solutions that would lead to improved health for individuals and populations”. The consortium’s eLwazi Open data Science Platform aims to address some of the aforementioned challenges to enable African scientists to access, share, analyze and interpret large, multidisciplinary datasets for novel health discoveries. This talk will describe the challenges and how the platform will address these in the context of African research settings.

• Martin Kelemen, University of Cambridge, United Kingdom
• Adam Butterworth, University of Cambridge, United Kingdom

Health disparities arise from variations in disease outcomes among different populations, influenced by both genetics and environmental factors. However, the impact of gene-environment interactions (GxE) at the country-level has remained an understudied area so far.
We relied on two large population cohorts of South Asian individuals in the UK and Bangladesh. Variance-component analyses and GxE-GWAS were performed on the following cardio-metabolic and anthropometric traits: height, BMI, years of education, systolic blood pressure (SBP), diastolic blood pressure (DBP), high cholesterol, myocardial infarction, stroke, type 2 diabetes (T2D) and angina.
We found significant Gene x Country variance-components for height (0.067), BMI (0.123), years of education (0.251), SBP (0.095) and DBP (0.080). Compared to single-cohort narrow-sense heritability estimates, we observed that for BMI, SBP and DBP, approximately half, and for years of education, close to the entire effect may be due to a country-level GxE term. However, among all traits, we found only a single significant locus (rs236554) for the T2D GxE GWAS.
Our results suggest that GxE effects may be common and likely to be due to a highly polygenic architecture comprising many SNP-environment interactions of small effect. Such GxE may exacerbate health inequalities by altering the efficacy of interventions across countries.

• Philip Bourne

Data science initiatives of some ilk are being developed in most academic research institutions worldwide. While their scope and emphasis may vary, I argue, they challenge the conventional way of thinking about bioinformatics, computational biology, and systems biology. It is more than a name change to become biomedical data science; it is an opportunity as I will explain. It addresses and updates the question I posed in 2021, Is Bioinformatics Dead? https://doi.org/10.1371/journal.pbio.3001165

• Guy Cochrane, Global Biodata Coalition, France

Progress in life and biomedical sciences depends absolutely on biodata resources - databases comprising biological data and services around those databases. Supporting scientists in data operations and spanning management, analysis and publication of newly generated data and access to pre-existing reference data, these biodata resources together comprise a critical infrastructure for the domain. Unlike other scientific infrastructures, biodata resources are globally distributed and lack any kind of central coordination. While this configuration supports innovation, it lends itself poorly to the long-term sustainability of individual biodata resources and the infrastructure as a whole. The Global Biodata Coalition (GBC) brings together life science research funding organisations that recognise these challenges and acknowledge the threat that the lack of sustainability poses. They agree to work together to find ways to improve sustainability.

In the presentation I will outline biodata resources and the infrastructure that they make up and discuss sustainability challenges. Covering some of the work that GBC has carried out to understand and classify biodata resources and the entire biodata resource infrastructure, I will outline the Global Core Biodata Resource programme and Inventory project. I will introduce the upcoming stakeholder consultation processes around approaches to sustainability and open data. Finally I will lay out the path GBC is taking to engage researchers, informaticians, funding organisations and other stakeholders in moving towards greater sustainability for these critical foundations for our field.

• Christophe Godin, INRIA, France
• Teva Vernoux, CNRS, France

• Philippe Sanseau

In this talk I will discuss how relevant data, especially genetics and genomics, alongside computational analysis are essential to uncover new biological insights as well as to identify and validate new drug targets.