Weekly Tech+Bio Highlights
The Rise of Subscription-based AI Tools: Analyzing CT Scans to Improve Clinical Trial Design; A New Platform for Polymorph Selection in Drug Development; I need your assistance (a short poll included)
Hi! I am Andrii Buvailo, and this is my weekly newsletter, ‘Where Tech Meets Bio,’ where I talk about technologies, breakthroughs, and great companies moving the biopharma industry forward.
If you've received it, then you either subscribed or someone forwarded it to you. If the latter is the case, subscribe by pressing this button:
Starting with this newsletter, I am testing a new approach where I would be sending two newsletters per week, on Sunday (free) and in the middle of the week (for paid readers).
The Sunday free post will mostly feature news highlights (like this one), while the mid-week post for paid subscribers will be more in-depth exploration, or a listicle with curated companies, technologies, and trends.
Please let me know in the comments what you think about this approach.
Now, let’s get to this week’s topics!
Weekly Tech+Bio Highlights:
🔬 XtalPi introduces XtalGazer, an AI-driven platform for polymorph selection in drug development, integrating AI, automation, and crystal structure prediction to enhance efficiency and accuracy in pharmaceutical research.
🔬 iBio and AstralBio announce a collaboration to develop novel antibodies for obesity and cardiometabolic diseases, leveraging iBio's AI/Machine Learning platform for drug discovery, alongside a $15M financing to support the partnership.
💰 Century Health secures $2M in funding to use AI for analyzing clinical data, aiming to enhance drug development and commercialization by providing pharmaceutical companies and researchers with access to valuable patient insights for diseases like Alzheimer’s.
🚀 Profluent, launched by Salesforce ProGen researcher Ali Madani, aims to revolutionize drug development using AI to design custom-fit proteins for therapies, backed by a recent $35 million funding round and supported by notable figures like Google's Jeff Dean.
The Rise of Subscription-based AI Tools
I’ve long been a fan of C&EN, since my early career days as a scientist (a chemist). Here is yet another interesting read, by Aayushi Pratap, The rise of subscription-based AI platforms for drug discovery.
The article discusses the emerging trend of subscription-based AI platforms in the drug discovery sector, highlighting the (claimed) democratization of access to AI tools for a broader range of researchers and companies.
Traditional drug discovery processes are expensive and time-consuming, prompting pharmaceutical giants to invest heavily in AI to expedite drug development. However, smaller biotech firms and research groups often find the costs prohibitive.
Software as a Service (SaaS) models are now making AI more accessible by offering cloud-based services at a monthly subscription fee. This development allows drug researchers to utilize AI for discovering new molecules without the need for extensive, personalized services that come with partnerships.
Key points from the article include:
Nvidia's Involvement: Nvidia, primarily known for its GPUs essential for AI operations, has introduced Nvidia Inference Microservices (NIM) as part of its AI suite for drug discovery, priced at $4,500 per GPU per year or $1.00 per GPU per hour. Read also my previous newsletter, where I reviewed microservices in more detail.
1910 Genetics' Approach: This Boston-based company uses its AI platform to develop drugs for neurological and autoimmune diseases and cancer. It has expanded its services through SaaS, allowing researchers to discover large-molecule medicines on its platform.
Cradle's SaaS Model: Cradle offers a generative AI platform for designing and engineering proteins via SaaS, making advanced machine learning tools accessible to smaller startups. In my recent interview, Dr. Elise de Reus, co-founder of Cradle, nicely explained what the company is doing.
While SaaS models are expanding access to AI for drug discovery, the industry is still navigating the balance between subscription services and partnerships with big pharmaceutical companies.
Some firms, like Atomic AI, choose not to offer SaaS, focusing instead on partnerships and in-house development for better drug approval odds.
AI-assisted CT Scans Analysis to Improve Clinical Trial Design
In a collaborative effort between Brainomix, an AI-driven medical spin-out from the University of Oxford, and AstraZeneca, a new study demonstrates significant advancements in the identification and stratification of patients with idiopathic pulmonary fibrosis (IPF) at risk of disease progression.
The research, published in the American Journal of Respiratory and Critical Care Medicine, leveraged Brainomix's e-Lung software, a cutting-edge artificial intelligence-enabled tool designed for automated processing of CT scans in clinical settings.
The study focused on analyzing data from AstraZeneca's Phase 2 clinical trial of tralokinumab, a treatment under investigation for IPF, utilizing the e-Lung platform to process patient data.
A notable aspect of e-Lung is its incorporation of the Weighted Reticulovascular Score (WRVS), an innovative biomarker that evaluates reticular opacities and vascular structures within the lungs, offering a novel approach to assessing disease progression risk.
Findings from this analysis revealed that WRVS, derived from a single baseline CT scan, could predict the decline in Forced Vital Capacity (FVC) over a 52-week period more effectively than traditional measures.
This capability suggests e-Lung's potential to enrich clinical trials with patients more likely to show disease progression, optimize trial design by ensuring well-matched treatment arms, and potentially reduce the overall size of future clinical trials.
This study builds on previous research presented at the American Thoracic Society meeting, which found WRVS to be a stronger predictor of transplant-free survival in IPF patients compared to FVC alone.
A New Platform for Polymorph Selection in Drug Development
A new AI-driven solid form discovery platform aims to improve the polymorph selection process for the pharmaceutical industry. XtalPi Inc., a US-China-based company integrating artificial intelligence and robotics, announced the launch of its proprietary comprehensive solid form discovery and selection platform, XtalGazer.
A key component of XtalGazer is XtalCSP, a crystal structure prediction platform to perform global searches of crystal structures for target molecules and the other optional components in the corresponding searching space, offering a deep insight into possible stable forms.
Furthermore, crystallization strategy recommendations will provide AI-backed experimental design to help avoid human bias. XtalGazer also utilizes MicroED to rapidly elucidate crystal structures from powder samples, reducing the need for growing single crystals.
This announcement is a nice example of how AI is applied beyond target discovery and drug design.
Feel free to write in the comments if there is a good idea that is not included in this poll.
Starting with this newsletter, I am testing a new approach where I would be sending two newsletters per week, on Sunday (free) and in the middle of the week (for paid readers).
The Sunday free post will mostly feature news highlights (like this one), while the mid-week post for paid subscribers will be more in-depth exploration, or a listicle with curated companies, technologies, and trends.