The hottest Drug Discovery Substack posts right now

And their main takeaways
Category
Top Science Topics

F Cancer

Marcus on AI • 13437 implied HN points • 16 Mar 26
🕹 Technology Drug Discovery
  1. Biology is incredibly complex and varies from person to person, so many drugs that look promising in animals or early tests still fail in humans.
  2. Current AI is not a magic cure—existing models are limited and often trained on language, so much stronger algorithms that can reason about chemistry, physics, and biology are needed for major breakthroughs.
  3. In the near term, AI can help by streamlining paperwork, patient recruitment, and researcher tools, but real progress also depends on economic and systemic changes like better incentives and funding.

Building a Lever on Time

New World Same Humans • 28 implied HN points • 22 Mar 26
🕹 Technology Drug Discovery
  1. World models can simulate physical reality and let us run thousands of virtual experiments in parallel, speeding up tasks like robot training, materials testing, and drug discovery.
  2. By turning compute and energy into synthetic time, these simulations can compress years of real-world processes into hours or minutes, acting as a powerful lever on time.
  3. The main challenge will be managing and interpreting the huge volume of simulated outcomes, so we’ll need better tools or machine assistance to surface useful insights and decide what to explore.

✨⚕️ Must AI cure cancer?

Faster, Please! • 1370 implied HN points • 24 Feb 26
🕹 Technology Drug Discovery
  1. AI doesn't have to instantly cure cancer to be a huge win. Even steady improvements that make treatments more precise and drug discovery cheaper would be transformative.
  2. AI is already helping reverse decades of falling pharma productivity by acting as a better front-end filter — boosting candidate success rates, shortening timelines by roughly 20–25%, and cutting development costs by about 25–30% — which could unlock tens to hundreds of billions in value.
  3. Apocalyptic job-loss stories are overstated because they ignore new job creation, the gap between lab capability and workplace adoption, and political and economic constraints that will slow large-scale disruption.

Weekly Dose of Optimism #180

Not Boring by Packy McCormick • 97 implied HN points • 13 Feb 26
🕹 Technology Drug Discovery
  1. AI drug design engines can now predict protein-ligand structures and binding strengths far faster and more accurately than older models, turning months of lab search into minutes of computation. If these predictions translate to real-world medicines, we could see many more novel drug candidates enter clinical pipelines, shifting bottlenecks to trials and regulation.
  2. New AI 'deep thinking' modes are able to spend minutes or longer reasoning through hard math, materials, and experimental problems, and can even generate lab-ready protocols for automated equipment. That capability points toward AI-assisted discovery and self-driving labs that amplify human researchers across disciplines.
  3. Researchers found a tiny 45-nucleotide ribozyme that can synthesize its complement and a copy of itself using trinucleotide building blocks, solving a major self-replication puzzle. Its simplicity makes a plausible origin-of-life pathway more likely, linking early replication chemistry to the genetic code we still use today.

LabGenius - the long slow road to faster drug discovery

Rory’s Always On Newsletter • 535 implied HN points • 07 Feb 24
🔬 Science Drug Discovery
  1. AI and machine learning are revolutionizing drug discovery by speeding up the identification of potential treatments, leading to big rewards for those in the industry.
  2. Building a successful biotech company requires patience, determination, and significant funding, often with a focus on research and development before revenue generation.
  3. Investors in biotech companies must be prepared for a long journey of constant failures and successes, akin to the process of drug discovery, with potential acquisitions being key outcomes.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

Cancer-Selective, Pan-Essential Targets from DepMap

Rough Diamonds • 20 implied HN points • 09 Jan 26
🔬 Science Drug Discovery
  1. Large-scale DepMap screening can flag genes whose knockout strongly harms many cancer cell lines but not immortalized "normal" lines, yet these results are limited by dataset definitions (many hits fall on DepMap's "pan-essential" list) and by the poor representation of healthy human tissues in culture, so experimental validation is needed.
  2. The top candidates include both familiar chemotherapy targets and new leads: some targets already have clinical-stage inhibitors or ADCs (e.g., TFRC, NMT1), while others (e.g., YRDC, SEPHS2, PHF5A, ADSL) are preclinical or underexplored and could be druggable by different modalities.
  3. LLM-generated code (Claude Code) made the project fast and reproducible, but agent-produced code can silently change behavior or omit checks, so careful human review, testing, and follow-up biological experiments are essential.

Measuring Drug Target Success

Rough Diamonds • 9 implied HN points • 16 Dec 25
🔬 Science Drug Discovery
  1. Most modern drugs are built around a specific molecular target, and researchers pick targets using genetic, animal, or in‑vitro evidence that suggests the target is causally involved in disease.
  2. Targets backed by human genetic evidence more than double a drug's chance of clinical success, while pursuing mechanisms similar to past failures increases the odds of failing.
  3. Preclinical signals can improve early selection but don't replace human trials, so improving the ROI of drug development means making trials cheaper and/or picking better candidates early, rather than relying only on rational design.

Nautilus Early Access

ASeq Newsletter • 14 implied HN points • 25 Nov 25
🔬 Science Drug Discovery
  1. Nautilus has been pushing an early-access program and that push seems to have increased market interest by showing the platform can support early-access projects.
  2. A recent scientific demo focused on Tau proteoforms (about 768), which is a useful small-scale result but doesn’t demonstrate the claimed ability to interrogate billions of wells or many different proteins.
  3. Because the demo was small, it’s unclear how well the high-density patterning and machine-learning pattern matching perform at scale, so fuller multi-protein or high-well-count demonstrations are needed.

Drug the Undruggable: Learn how Elsie Biotechnologies is advancing oligonucleotide therapeutics

LatchBio • 268 implied HN points • 07 Mar 24
🔬 Science Drug Discovery
  1. Elsie Biotechnologies uses computational design tools and high-throughput experimental approaches to develop drugs.
  2. Partnerships with large pharmaceutical companies like GSK can significantly impact the progress of smaller biotech companies.
  3. Oligonucleotide therapies represent the future of drug development, promising safer and more effective treatments.

Vial: Reimagining Clinical Trials

The Century of Biology • 308 implied HN points • 30 Jul 23
🏥 Health & Wellness Drug Discovery
  1. Vial is working to reduce the cost of clinical trials by innovating new technology.
  2. Vial's Act I focuses on structuring, digitizing, and automating the clinical trial process to drive efficiency.
  3. Vial's Act II, Battery Bio, aims to revolutionize drug discovery by integrating software and advanced technologies in a vertically integrated approach.

Gradient Ascent #14

Gradient Ascent • 19 implied HN points • 20 Mar 23
🔬 Science Drug Discovery
  1. Natural compounds from nature are a valuable source for potential new drugs.
  2. Graph neural networks are used to predict mass spectra from molecular structure in a novel way.
  3. The graph neural network approach provides a promising path for computational drug discovery advancements.

Unbiased mapping of cereblon neosubstrate landscape by high-throughput proteomics

Axial • 14 implied HN points • 24 Nov 24
🔬 Science Drug Discovery
  1. A new method helps find powerful compounds that can target hard-to-reach proteins for drug development. These compounds are called molecular glue degraders, and they can help break down unwanted proteins in the body.
  2. The study found many new targets for these compounds, including some that haven't been studied much before. This expands the potential for developing new treatments for diseases like cancer.
  3. The researchers created a process that combines different scientific techniques, making it easier to design and improve these drugs. This means we might see more precise and effective medicines in the future.

Screenshots to Code Dataset, Multi Motion Brush in Gen-2, Open-source AGI, AI system that solves complex geometry problems, AI in drug discovery and more

AI Brews • 22 implied HN points • 19 Jan 24
🕹 Technology Drug Discovery
  1. Google DeepMind's AlphaGeometry AI system solves complex geometry problems at human Olympiad level.
  2. Codium AI's AlphaCodium improves code generation in LLMs with test-based iterative flow.
  3. Meta is working on open-source AGI and Microsoft Research made progress in AI-driven drug discovery.

DNA foundation modeling

Axial • 7 implied HN points • 10 Dec 24
🔬 Science Drug Discovery
  1. DNA foundation models are helping scientists analyze and understand the complex patterns in genetic data. They can lead to important discoveries in medicine and biology.
  2. Building these models is tough because DNA sequences are long and complicated. Special techniques are needed to process them efficiently and recognize important details.
  3. While these models have great potential, they need to be tested carefully to avoid mistakes. We also need to think about the ethical implications of using them in research and medicine.

Axial Discovery - Phenotypic screening

Discovery by Axial • 3 implied HN points • 27 Mar 23
🔬 Science Drug Discovery
  1. Phenotypic screening focuses on identifying specific physical or biochemical traits of interest for drug discovery.
  2. Key rules for effective phenotypic screens include selecting relevant cell models, designing disease-specific assays, and defining clinical-like endpoints.
  3. Advancing phenotypic screening requires improving throughput of complex models, developing translational disease models, enhancing proteomic tools, and integrating phenotypic and target-based screening.

Unfolding AlphaFold 3 🧬✨

Sector 6 | The Newsletter of AIM • 0 implied HN points • 11 May 24
🕹 Technology Drug Discovery
  1. AlphaFold 3 is an advanced AI model that improves protein and molecule interaction predictions by 50%.
  2. This technology goes beyond just analyzing protein structures to help design drug compounds that can bind to proteins.
  3. The goal of this AI is to enhance drug discovery, making it easier to create effective treatments.