@PheironAI

We underwrite biology risk to engineer Biological Alpha: knowing what biology will work in humans early enough to change the outcome of a drug program.

 New York
pheiron.ai
Joined October 2023
  • Tweets80
  • Following4
  • Followers266
21Media
On September 23, Remo Monti will take the stage at Target Validation - In the Era of Genomics, Big Data, and AI. His talk will focus on one of the central questions in drug discovery: which sources of evidence actually predict clinical success?
1
86
Remo will share how Pheiron evaluates a vast number of biological evidence sources and how agentic workflows can turn that complexity into better drug-target decisions.
1
63
Predictive validity of evidence for drug-target selection at scale - moving beyond genetics September 23, 12:45 Wellcome Genome Campus, Cambridge If you’re attending and want to learn more about how Pheiron measures biological risk, find Remo and say hello.
28
Novartis Phase 3 Lp(a)HORIZON trial read out on 4 September 2026 and did not meet its primary endpoint. The drug substantially lowered Lp(a), but the topline didn’t show a statistically significant reduction in cardiovascular events.
1
1
1
144
more replies
The risks in this trial were how much of the lifetime genetic effect would translate within four years, and whether the ~70% sustained Lp(a) lowering we modelled for monthly pelacarsen would be enough to add benefit in a population receiving intensive cardiovascular treatment.
1
40
A prediction is useful when it also shows what drives the probability. In this case, we identified and quantified the biological and trial-level risks before the readout. The full prediction is here: pheiron.com/predictions/nct0…
24
New podcast out today: @buergelt_ x @DrSamuelBHume They cover: > why human qualitative thinking limits drug development > what it takes to quantify biological risk > why we publish predictions on upcoming clinical trials before they read out
1
4
163
> what happened so far: we got most right, and missed one > why “more data” ≠ immediately better decisions in drug development Link for the full podcast: youtube.com/watch?v=oU1ui60B…

Thore Buergel, CEO at Pheiron, on predicting clinical trials with AI

Dr. Thore Buergel is CEO at Pheiron, a company dedicated to predict...

youtube.com
5
93
Better models alone won’t solve drug discovery. We need better infrastructure.
1
3
140
more replies
We also gave the group an early look at our blog, The Engineering behind Biological Alpha, our perspective on why progress requires more than increasingly capable models. Thank you to everyone who came and joined us.
1
50
If you missed the blog, you can find it here: blog.pheiron.com/p/the-engin…

The Engineering behind Biological Alpha

At the current pace of drug discovery, it will take more than 500 years to develop treatments for every known disease.

blog.pheiron.com
23
Drug discovery depends on information. But biological data, software, and compute are fundamentally disjoint. Before information can drive a decision, it first has to be engineered.
1
139
Human scientists waste lots of time doing exactly that and today's agents fail when released from constrained environments. We formalize the cost of engineering information from disjoint data, software and compute as Information Impedance.
1
69
To truly enable agentic science, we have to build infrastructure that reduces Information Impedance. Our latest blog, The Engineering Behind Biological Alpha, explains how we’re solving it. Link: open.substack.com/pub/pheiro…
24
Nine out of ten drug programs entering clinical translation fail. Tons of assays and experimental methods producing data on biology are available at the pre-clinical stage.
1
1
5
335
more replies
In our latest blog, we walk through how benchmarking data is the foundation for decision making at Pheiron.
1
72
Link to our substack: blog.pheiron.com/p/measuring…

Measuring Biological Alpha

Benchmarking expected P(Success) lift from preclinical data sources.

blog.pheiron.com
65
Load more