About the PULSE program

This area integrates the disciplines of genetics and biochemistry with the ambition to fully understand how cells, the fundamental unit of life, function.  Progress is made by analyzing the molecular components in time and space, from single molecules to native tissue environments at high resolution. This research area is rapidly evolving with the development and application of novel high-density data driven methods, often relying on machine learning, artificial intelligence, or other computational techniques to analyze, integrate and interpret large sets of imaging data. The ultimate goal is to create predictive models for core functions of cells and tissues.

At SciLifeLab, research in this area is aligned with the EU definition of precision medicine, ie, “characterization of individuals’ phenotypes and genotypes to tailor therapeutic strategies, assess disease predisposition, and enable timely, targeted prevention.” Accordingly, here the aim is to integrate molecular and clinical data to identify biomarkers for disease risk and diagnosis, evaluate drug responses, and support health monitoring. Research projects are expected to contribute with strong capabilities inmachine learning, AI and other computational tools. Possibilities exist to mine strong assets in Sweden, such as molecular data (e.g. omics), imaging, electronic healthcarerecords, longitudinal patient and population registries, biobanks and digital monitoring data.

This subject area concerns research using large experimental, clinical, or pathogen surveillance data in innovative ways to transform our understanding of pathogens, their interactions with hosts and the environment, and how they are transmitted through populations. The aim is to meticulously analyze pathogen-host model systems, exploiting multidimensional and genome-scale experimental data that are now routinely attainable, and extending to population-scale genetic, clinical, or public health data from pathogen surveillance efforts and biobanks. The ultimate goal is to enhance the ability to anticipate, prevent, and respond to infectious disease outbreaks, ensuring a proactive approach to safeguarding public health.

Projects within this area address life from a trans-disciplinary perspective, encompassing molecules and cells to ecosystems. Researchers are leveraging the substantial data streams generatedby high-throughput sequencing of genomes and biomes,continuous recording of video and audio in the wild, high-throughput imaging of biological specimens, and large-scale remote monitoring of organisms and habitats. Novel insights are likely to emerge by exploiting the development and applicationof novel methods relying on machine learning, artificial intelligence, or other computational techniques, enhancing our understanding of population genetics, adaptation, speciation, convergent evolution, and functional constraints. A deeper understanding of the complex interactions within ecosystems and their implications for planetary health is needed to find solutions to impede the ongoing loss of biodiversity.

This is an exceptional opportunity to conduct research at the intersection of artificial intelligence (AI)/Machine Learning (ML), data science, and drug discovery. The fellow will lead the design, development, and implementation of novel AI/ML methods for preclinical drug discovery of small molecules, proteins, or oligonucleotides, focusing on developing and applying AI/ML methods in hit-to-lead optimization projects for any of these drug modalities.

Key project areas include, but are not limited to, the development of AI/ML tools, web-based workflows, and novel approaches in deep learning, de novo design and generative AI for drug discovery. Candidates are encouraged to apply AI/ML methodologies across preclinical drug discovery project based on small molecules, proteins, oligonucleotides, or emerging “new modalities”. Examples of areas of research could be: novel molecular representations and generation, chemical space navigation, ultra-large virtual screening, multi-parameter optimization, ligandability assessment, protein design, RNA and oligonucleotide structure prediction, prediction of on/off-target oligonucleotide binding and effect, computational characterization of enzymatic RNA cleavage, antibody developability, and de novo design. We also welcome proposals from emerging fields in drug discovery, including explainable AI, and integrating ‘first-principles’ with AI/ML (e.g., quantum machine learning approaches).

Examples of challenges in this area:
•There is very limited access to high-quality, standardized datasets, combined with imbalanced data (e.g., active vs. inactive compounds) which hampers model training and generalizability. How to access these data sets?
•Many models, particularly deep learning-based ones, suffer from a lack of interpretability, making it difficult to validate predictions and gain insights into molecular mechanisms. Methods to solve this problem?
•Scalability is another barrier, as handling large datasets or running complex simulations requires significant computational resources. How to speed up things? 
•The interdisciplinary knowledge gap between AI researchers and domain experts in chemistry and biology slowsprogress, as effective collaboration requires bridging expertise across these fields. Can these barriers be torned down?

The successful candidate will join a vibrant, multidisciplinary research environment, collaborating with bioinformaticians, structural biologists, chemists, and drug discovery experts.

Contact: vasanthanathan.poongavanam@scilifelab.uu.se

Through the OligoNova Hub units in Gothenburg, SciLifeLab-DDD offers state-of-the-art capabilities for oligonucleotide-based drug discovery, including but not limited to antisense and siRNA-based therapeutic modalities, as well as modalities that conjugate oligonucleotides to SciLifeLab-DDD’s other modalities such as proteins and small molecular ligands. In-house capacity for bioinformatics and AI/ML-guided oligo design, oligonucleotide synthesis, automated and high-throughput analyses of cellular effects through qPCR, NGS, mass spectrometry, imaging, and other techniques is complemented with advanced proteomics, NMR, and additional imaging modalities including Nano-SIMS within the local core facilities organization and other collaborators.

New technologies to design and analyse therapeutic oligonucleotides to support the development of optimally efficacious and safe therapeutic agents are continuously developed and applied within the focus area. Current examples include the development of a modular reporter assay system to evaluate target knockdown, new computational techniques for nucleic acid structure prediction, and biophysical technologies for the characterization of oligonucleotide binding interactions.

Of great interest are new technologies and approaches that could streamline any aspect of the discovery process, provide new molecular and/or mechanistic insight into the action and safety of therapeutic oligonucleotides, provide new means of measuring oligonucleotide distribution, exposure, and effects in vitro and in vivo, or enable the design of novel tissue- and cell type-selective oligonucleotide delivery strategies. In addition to experimentally oriented projects, AI/ML-approaches or other computational techniques could be integrated project components.

Examples of challenges in this area:
•New methods to characterize and prioritize therapeutic oligonucleotides in large libraries with regard to efficacy and/or safety
•New methods to measure tissue- and (sub)cellular distribution of oligonucleotide therapeutics
•New strategies for enhancing delivery of oligonucleotide therapeutics to specific tissues and cell types through conjugation to targeting molecules, and new methodologies to identify and characterize such targeting strategies
•Methods for better prediction of stability and  tolerability/safety of therapeutic oligonucleotides

Contact: par.matsson@gu.se

Within the SciLifeLab Display and Selection Technologies focus area we have access to state-of-the-art platforms for discovery of therapeutic antibodies, small molecules and oligonucleotides. These include e.g. phage display libraries as source of human antibodies, DNA-encoded chemical libraries (DELs) for small molecules, automated systems to facilitate high-throughput screening and instruments to study binding of therapeutic candidates to proteins and cells. Through close collaborations with other SciLifeLab units, local university networks and CROs additional instruments can be accessed e.g. mass spectrometry, DNA sequencing (Sanger, NGS, long-read) and various imaging techniques.

Technology development is continuously ongoing within the focus area. Currently a single-domain antibody library in phage display format as well as new DELs are under construction, antibody tools for targeted protein degradation and intracellular delivery are developed, new reporter systems for screening of oligonucleotides are set up and long-read sequencing is explored as a complement to traditional antibody screening.

We welcome technology development proposals improving any step of the early drug discovery process. Areas of interest could include e.g. novel selection and screening strategies, alternative display technologies, new therapeutic antibody formats, strategies for optimization of binding affinity and strategies for improved developability. AI-driven approaches and NGS/long-read sequencing could be integrated elements in such development work.

Examples of challenges in this area:
•Presently, recombinant protein displayed on beads are used for DEL screening, Can a technology be developed that allows for in situ selection from large DELs in human cells or bacteria?
•Can NGS-based strategies be developed to bypass needs for screening the output from selections in phage display libraries?
•Engineering of candidate drug antibodies for preclinical development is circumstantial. Can new strategies be developed to make human antibody maturation and the optimization of developability aspects more efficacious?
•Novel technologies and vectors that generate a better therapeutic window for CAR applications are in demand. How can this be accomplished?
•Development of novel display/array formats for antibody/peptide screening and development

Contact: mikael.mattsson@immun.lth.se

Proximity-Inducing Agents are an emerging approach of therapeutics that leverage the ability to bring two or more biomolecules into proximity to induce specific biological effects. These agents are based on technologies such as bi-specific antibodies, antibody-drug conjugates, or various small molecule approaches for intracellular re-wiring of protein localization. Emerging technologies for Targeted Protein Degradation (TPD), such as PROTACs (Proteolysis Targeting Chimeras) and AbTACs (antibody-based targeted protein degradation agents), show promise in precisely targeting and degrading disease-causing proteins, potentially offering new treatment avenues that conventional drugs may not address.

We offer exceptional opportunities to conduct research that will shape the technological advancement of proximity inducing therapeutics based on modalities that is compatible with areas within scope for the SciLifeLab DDD platform, i.e. small molecules, proteins, oligonucleotides, or new modalities. Cell-based therapies are out-of-scope.

Depending on the candidate’s background and interests we support proposals related, but not limited, to: small molecule degraders, antibody-based degraders for extra-cellular or membrane proteins, targeted delivery of payloads (oligonucleotides, small molecules, radioligands, degraders), bi-specific antibodies, etc. Currently, we have toolboxes, assays, and methodology for small molecule- or antibody-based degraders and targeted delivery available. We offer industry standard facilities and infrastructures for e.g. organic synthesis – including oligonucleotides, biochemical and cellular assays, biophysics, ADME, protein expression, phage-display selections, computational and machine learning capabilities, etc. SciLifeLab offers opportunities to access additional world-class infrastructures.

Examples of challenges in this area:
•New methods that enable the identification of bifunctional compounds and molecular glues that act on pre-selected presenter and target proteins, e.g. by comparing barcode enrichment after screening DNA-encoded libraries in the presence or absence of presenter proteins
•Novel antibody formats with applications for targeted protein degradation
•Innovative technologies that identifies drug-like molecules with high affinity to unstructured proteins

Contact: Per.Arvidsson@scilifelab.se