Right from finding potential active substances as well as identifying novel targets to simulating how drugs will perform within the human body and also optimizing laboratory workflows, digital technologies are creating a massive impact when it comes to drug discovery and development. Technologies such as AI and ML, as well as robotics and automation, are helping researchers process huge data sets and also conduct preclinical investigations in historic time, thereby speeding up the process and also offering greater precision.
But many have cautioned against too great a dependence on these tools and have emphasized their significance when it comes to checks and balances so as to ensure the systems happen to be properly regulated and the results are verified.
Companies that are leading the way
There are numerous companies leading the charge when it comes to applying innovative technologies, especially AI and ML, when it comes to the search for new medicines.
Seven AI-focused companies have been highlighted by Clarivate in its 2023 ‘Companies to Watch Report’1:
- AQEMIA, which makes use of quantum and statistical mechanics algorithms so as to fuel generative AI to design novel drug candidates.
- Auransa, which harnesses the available data in order to discover highly intractable diseases having a poorly understood biology.
- Enveda Biosciences leverages a blend of ML, metabolomics as well as automation in order to identify bioactive, plant-based molecules of interest.
- Pharos iBio, which makes use of AI technology so as to develop treatments for refractory and rare diseases like cancer and repurpose existing therapies.
- Quris-AI, which mixes patient-on-a-chip AI and real-time nano-sensing technologies to forecast the safety as well as efficacy of drug candidates.
- Relation Therapeutics that makes use of high-resolution biology, ML, along with clinical insights to discover transformational medicines.
- SandboxAQ, which mixes AI as well as quantum technology- AQ in order to advance drug discovery for intricate, undruggable targets.
Companies such as these, which can provide digital technology insights, are not to be surprised, enjoying considerable popularity. Biopharma as well as biotech companies’ sans such capabilities are keen to make the utmost use of these technologies, and the number of strategic collabs with such companies happens to be accelerating at an unprecedented rate. Sanofi has gone ahead and stated that it looks to become the first pharma company that is powered by AI at scale, and as part of this endevor, has agreed to collaborate with BioMap so as to co-develop advanced AI models and protein Large Language Models, which it hopes will go ahead and help biologics design and multiparametric optimisation2.
In 2024, Amgen went ahead and announced its plans to make use of NVIDIA’s DGX SuperPOD-powered insights by way of human datasets to create a human diversity atlas when it comes to drug target and biomarker discovery. VantAI as well as Blueprint Medicines also extended their existing collaboration in order to focus on undruggable targets by making use of induced proximity drug discovery.
In yet another early 2024 deal, one-to-watch SandboxAQ went on to acquire Good Chemistry, a computational chemistry organization that uses AI, quantum, along with other advanced technologies so as to accelerate medicine research3.
Arman Zaribafiyan, Good Chemistry founder and CEO, remarks on the significance of these new high-tech resources, wherein he states that the quick advances in AI, quantum, cloud and high-performance computing have gone on to unlock endless opportunities for firms such as Good Chemistry and SandboxAQ so as to reinvent the way they think pertaining to chemistry, discover ways to have the products safer, more robust, and at the same time more sustainable, and also reshape the fabric of the world.
Universities are also making use of these digital technologies in order to advance their research. Late in 2023, SimBioSys as well as the Chan Lab at UT Southwestern went on to reveal plans in order to enhance PhenoScope’s tumor bank by way of spatial transcriptomic data with the objective of coming up with cancer biomarkers for immunotherapy.
Speeding up drug discovery
In turn, this guidance as well as investment when it comes to new technology is leading to a meaningful impact when it comes to the speed and efficiency of drug discovery as well as development.
One barrier when it comes to drug discovery is coming up with new chemical structures that have drug-like traits, and this is a region in which AI comes to the rescue. A biotech company such as Gero uses a hybrid quantum-classical machine-learning model so as to interface between classical and quantum computational devices in order to generate novel chemical structures when it comes to potential drugs. It went right, with the model suggesting that 2,331 novel chemical structures have properties typical of biologically active compounds, less than 1% of which happen to have a high similarity to any molecule in the training set4.
The CEO of Gero, Peter Fedichev, says that these breakthroughs go on to pave the way for a dramatic speeding of the drug discovery process. Drug design functions at the intersection of the classical and quantum phenomena realms and needs simultaneous determination when it comes to the quantum properties of drug-like molecules as well as their effects on living systems, as described by classical physics. This is where quantum computing will massively push the capacity to create transformative treatments for the most challenging diseases as well as conditions such as aging itself.
Digital tech is also aiding researchers when it comes to finding target candidates for some of the steepest disease areas. University of Oslo researchers, as well as researchers from the University of Chicago Pritzker School of Medicine as well as Insilico Medicine, made use of an AI target discovery engine so as to analyze transcriptomic data in order to identify dual targets when it comes to cancer and aging5. The research went on to identify a number of potential age-associated cancer targets. The team was able to validate a candidate that looked the most promising, and that was the gene histone demethylase, or KDM1A.
Assistant Professor of Medicine at UChicago in the section of hematology and oncology, Evgeny Izumchenko, said that they were very encouraged by the results, as this is a first study that showed the feasibility of AI-driven approaches in order to identify potential dual-purpose targets for anti-aging as well as anti-cancer treatment and goes on to absolutely demonstrate the value of such tools when it comes to addressing the intricate challenges at the interface between aging as well as carcinogenesis.
They are indeed beginning to witness a visible impact when it comes to new technologies in the sector, since drugs discovered by way of the use of AI start so as to enter clinical investigations. Verge Genomics’ amyotrophic lateral sclerosis- ALS treatment, VRG50635, holds the potential to be one of the first drugs to enter clinical trials that was completely discovered and developed by making use of an AI-enabled platform6. The company’s platform assessed over 11.4 million data points from ALS patient tissue as well as genetic datasets in order to discover loss of endolysosomal function as a novel causative mechanism when it comes to ALS and uncovered PIKfyve as a new therapeutic target that looked promising. VRG50635 has so far gone on to enhance the survival of ALS patient neurons and, at the same time, has also shown efficacy in a range of preclinical studies in ALS-relevant models of motor neuron degeneration.
Yet another AI-powered drug discovery as well as development company, BPGbio, has gone on to move to Phase II development when it comes to its lead candidate for advanced pancreatic cancer treatment, BPM31510. Thus far, the results go on to show BPM31510 happened to be well-tolerated, doubled progression-free survival vs. standalone chemotherapy, and have gone on to prompt more investigation of BPM31510 as a first-line therapy.
Taking on antimicrobial resistance
Antimicrobial resistance- AMR or antibiotic resistance, happens to be one of the greatest threats that human health faces today, and digital technologies are playing a role in this spectrum as well.
One company that is making use of AI models so as to tackle the issue when it comes to AMR is Obulytix, which is a spin-off based on research inferences from Ghent University as well as KU Leuven7. The company has gone on to come up with a novel AI platform in order to develop enzymes from bacteria-killing viruses- bacteriophages in a novel way so as to tackle bacterial infections, and has gone on to secure a €4 million investment in order to support its development.
Obulytix is not the only one. In late 2023, Massachusetts Institute of Technology- MIT researchers went ahead and revealed that they had gone on to identify a new class of antibiotic candidates against gram-negative bacteria by way of using deep learning8. The freshly discovered compounds can go on to kill methicillin-resistant Staphylococcus aureus- MRSA that are grown in a lab dish and in two mouse models of MRSA infection. Research on the AI model itself went on to offer the team more insights into how to track potential antibiotics. The researchers were able to gauge what kinds of data the deep-learning model was making use of so as to make its antibiotic potency predictions, which could go on to aid the researchers in designing additional drugs that may as well work even better.
The robotic revolution
Once targets as well as the molecules have gone on to be identified and simulations run, the next step within the process requires a move to the laboratory so as to validate the candidate. This is where robotics as well as automation are making a difference, often in conjunction with AI models.
In June 2023, Opentrons, which happens to be a lab automation company, went ahead and launched its Flex robots for scientists that were using generative AI. Its CEO of operations, Jon Brennan-Badal, says that it has been far too long, that the scientists have been constrained by their laboratory tools, and that by making lab automation as easy as a smartphone, the Flex robot democratises the access to automation within life sciences research.
Automation also happens to be applied to genomics sample preparation, which is an area that needs particular robustness as well as scalability. The Advanced Sequencing Facility at the Francis Crick Institute happens to be working with the company Automata in order to automate workflows, with the objective of reducing manual touchpoints, accelerating throughput, and making the utmost use of R&D flexibility. The walkaway workflows enable the researchers at the facility to prepare samples and also generate data rapidly, such as the validation of CRISPR genome editing, the extraction when it comes to genetic material from tumour samples, and genomic surveillance of the COVID-19 pandemic10.
Regulating digital technology
In spite of the breakthroughs as a result of such new insights, some have asked to exercise caution, especially in relation to a greater reliance on AI as well as ML. The European Medicines Agency- EMA published draft guidelines on the usage of AI in July 2023, underscoring that a human-centric approach should go ahead and guide all development as well as deployment of such technologies12.
The EMA report goes on to specify that AI and ML tools could be made use of to replace the usage of animal models in terms of preclinical development, so as to support the selection of patients for clinical trials or even draft, compile, translate, and review data. This breadth of applications goes on to bring a new set of challenges, like the understanding of possible biases within the algorithms, the challenge of technical failures, as well as the wider effect these would have on AI uptake when it comes to medicine development.
Data Analytics Centre at the Danish Medicines Agency Director and also the co-chair of the joint HMA-EMA Big Data Steering Group- BDSG, Jesper Kjær, says that AI use is quickly developing in society, and as regulators, they see more applications in the field of medicines. AI goes on to get exciting opportunities in order to generate new insights and also improve processes. To embrace them completely, they will have to be prepared for the regulatory issues presented by this rapidly evolving ecosystem.
Biosecurity happens to be an issue that has come into existence more recently because of the development of these new technologies, specifically biodesign tools and specialized AI models that happen to be trained on biological data and also offer insight into biological systems. The Federation of American Scientists- FAS has gone on to publish recommendations in order to address the requirement for oversight of biodesign AI tools, biosecurity screening of synthetic DNA, as well as guidance on biosecurity practices for automated laboratories13.
The recommendations go on to include institutional oversight for such tools from the government, the necessity to establish standards for assessing their risks, broadened infrastructure for cloud-based computational resources, biosecurity screening when it comes to synthetic DNA, as well as government guidance when it comes to biosecurity practices for automated laboratories.
FAS says that AI is most likely to furnish tremendous advances when it comes to the basic understanding of biological systems and significant benefits for health, agriculture, as well as the broader bioeconomy. But AI tools, if not used the way they should be or developed irresponsibly, can also go on to pose risks to biosecurity. The biosecurity spectrum of risks related to AI happens to be complex and also quickly changing, and gauging the range of issues needs diverse perspectives as well as expertise.