Slide 1: Recursion L(earnings) Call
o- Recursion Learnings) Call
Slide 2: Disclaimers
Disclaimers This presentation and any accompanying discussion and documents contain information that includes or is based upon "forward-looking statements" within the meaning of the Securities Litigation Reform Act of 1995. These forward-looking statements are based on our current expectations, esti
Slide 3: What is L(earnings) and why are we starting this practice now?
What is L(earnings) and why are we starting this practice now? Traditional Earnings Scripted, Boring, Hard to access L(earnings) Authentic, Adaptive, Easy to access 3 vs.
Slide 4: TechBio Origins - One Decade Ago
TechBio Origins - One Decade Ago 4
Slide 5: There is a formula for mapping and navigating complex systems using technology
There is a formula for mapping and navigating complex systems using technology Digital (bits) Data Aggregate and organize data to create digital maps of reality 2 2 Real (atoms) 1 Profile Systems Capture high-dimensional data to create a digital record of reality (things, places, preferences, etc.)
Slide 6: Data roadblocks made mapping and navigating biology difficult
Data roadblocks made mapping and navigating biology difficult Reproducibility Crisis Multiple studies have shown that the vast majority of published academic literature cannot be recapitulated Analog Standard The fax machine is alive and well in medicine, while in biopharma, study results from CROs
Slide 7: AI Expert Systems given way to Modern AI
AI Expert Systems given way to Modern AI Automation Automation tools enable massive scale Storage 1M-Fold Decrease In Costs over 40 years Compute 1M-Fold Increase in Compute over 40 years Bio Tools Tools like CRISPR allow CONTROL of Biology Why was the early-2010s the time for a step- function chang
Slide 8: Fast Forward to Today
Fast Forward to Today 8
Slide 9: Recursion OS: data, profile systems, algorithms
Data Each week we digitize millions of our own experiments across multiple layers of biology from cell to animal Improved and scaled clinical pipeline We are building and aggregating the right datasets to map and navigate biology Recursion OS 9 Algorithms We own and operate one of the fastest superc
Slide 10: The Recursion OS combines many tools to industrialize drug discovery
The Recursion OS combines many tools to industrialize drug discovery 10
Slide 11: Leading in TechBio in 2024
PIPELINE • First Generation: Five Phase 2 programs enrolling or soon to enroll patients focused in niche rare disease indications • Second Generation: Multiple preclinical programs and dozens of discovery stage programs focused in precision oncology PARTNERSHIPS • Bio: Large discovery collaborations
Slide 12: 2023 Year in Review
2023 Year in Review 12
Slide 13: May – Platforms: Acquisitions bolster digital chemistry and generative AI
(Toronto) (Montréal) • Enhance the optimization of Recursion’s compounds for efficacy while minimizing liabilities • Rapidly advance the diversification and discovery of novel chemical matter • Enables mechanism of action deconvolution and generative chemistry • Enable acceleration of generative des
Slide 14: June – Pipeline: REC-994 for CCM Phase 2 Completed Enrollment
June – Pipeline: REC-994 for CCM Phase 2 Completed Enrollment • First therapeutic candidate advanced to an industry-sponsored Phase 2 trial (SYCAMORE) for CCM • Fully enrolled ahead of schedule in June 2023 with 62 patients across 3 arms in a 1:1:1 randomization • Majority of patients treated with R
Slide 15: July – Partnership: Our NVIDIA Partnership Announcement
Trademarks are the property of their respective owners and used for informational purposes only. • Partnership on advanced computation (e.g., foundation model development) • Priority access to compute hardware or DGXCloud Resources “If we apply the same methodology that we use in computer-aided chip
Slide 16: August – Platform: Computation at scale with Enamine and MatchMaker
~36 Billion Compounds from the Enamine Real Space ~80,000 predicted binding pockets from ~15,000 human proteins ~2.8 Quadrillion potential protein-ligand interactions computed and stored Recursion partnered with to integrate and optimize MatchMaker (ac quired via ) for massive scale GPU-based comput
Slide 17: September – Pipeline: Phase 1 Study for REC-3964 Complete
September – Pipeline: Phase 1 Study for REC-3964 Complete • REC-3964 was safe and well tolerated at multiple doses up to 900 mg • No SAEs observed and no discontinuations related to treatment • Favorable PK profile with exposures (AUC) increasing approximately dose-proportionally • REC-3964 exposur
Slide 18: September – Platform: Recursion built Phenom-1
The combination of scaled data generation and accelerated computing is a key to advancing biological ML 18 September – Platform: Recursion built Phenom-1, the world’s largest phenomic foundation model
Slide 19: November – Platform: BioHive expansion
Computational Tools Expand BioHive-1 from: • 320 NVIDIA A100s… …to include an additional • 504 NVIDIA H100s With operations beginning H1 2024 Likely to be the highest performing compute cluster owned and operated by any biopharma company on earth and among the top 50 compute clusters on the Top500 l
Slide 20: October – Partnership: Roche exercises its option
• Roche exercises its Small Molecule Validated Program Option • Hit series identified using fit-for-purpose oncology map • Recursion will continue to take the lead in advancing the program, leveraging the Recursion OS and its suite of digital chemistry tools with the support and collaboration of Roc
Slide 21: November – Partnership: Recursion partners with Tempus
+ • $160M paid by Recursion to Tempus in cash or equity, at our election, in increasing annual increments over five years, beginning with $22M of equity to be issued later this year • Expected to accelerate model deployment, linking molecular data with outcomes • Expected to enhance program translat
Slide 22: November – Partnership: Update of existing Bayer collaboration
“The methodology in which Recursion uses artificial intelligence (AI) in drug discovery, could be one of the most disruptive technologies of our time... As our collaboration and the usage of AI continue to evolve, we look forward to continuing to work with industry innovators to identify novel targe
Slide 23: Next Steps: REC-1169575 fibrotic disease program
Next StepsFurther Confidence Reversal of Fibrocyte Differentiation Assay Identify a potential first-in-class therapeutic NCE with a novel MOA capable of reversing disease-related fibrotic processes Recursion-generated hits show concentration-dependent rescue in a disease relevant human PBMC assay an
Slide 24: December – Platform: Over 1 trillion hiPSC-derived neuronal cells produced
Q4 – Neuro iPSC production slide? Worlds largest producer 24 December – Platform: Over 1 trillion hiPSC- derived neuronal cells produced since 2022
Slide 25: More than a dozen discovery and research programs
25 More than a dozen discovery and research programs in oncology or with our partners All populations defined above are US and EU5 incidence unless otherwise noted. EU5 is defined as France, Germany, Italy, Spain, and UK. (1) Prevalence for hereditary and sporadic symptomatic population. (2) Annual
Slide 26: The Future of TechBio
The Future of TechBio 26
Slide 27: TechBio Origins: Point Solutions
TechBio Origins: Point Solutions Most BioTech companies have built a point solution - they’ve developed a tool, process, model or analysis to accomplish an important step in drug discovery. This is how we started too. But discovering and developing medicines requires hundreds of steps… 27
Slide 28: As point solutions evolve they increase in complexity and scale
As these point solutions evolve they increase in complexity and scale We manipulate human cells with CRISPR/Cas9- mediated gene knockouts, compounds, and other reagents Phenom-1 Groundbreaking models trained on >1 billion images and hundreds of millions of parameters learn to extract biologically me
Slide 29: To industrialize drug discovery, point solutions must be integrated as modules
To truly industrialize drug discovery, point solutions must be integrated as modules across many diverse steps 29
Slide 30: Each module is complex, and we continuously improve them
Each module is complex, and we continuously improve them Prioritize compound synthesis for compounds predicted to have high likelihood of suitable pharmacokinetics DMPK In Vivo Validation Establish in vitro-in vivo and in silico-in vivo correlations to minimize experimental toil. ANIMAL PHARMACOKINE
Slide 31: Utilizing each module requires specialized teams and expertise
Utilizing each module requires specialized teams and expertise InVivomic prioritization Digitized data collection yields real-time, continuous, and non-invasive data recorded in the animal’s home cage. Data generates high-dimensional assays, and ML connects studies for productivity. Overall, there i
Slide 32: We continuously add new modules to improve the Recursion OS
We continuously add new modules to improve the Recursion OS The Tempus data is computed and ML oncology models run on BioHive-1, Recursion’s in-house supercomputer. BioHive-1 will be expanded into a top 50 supercomputer in 2024 in partnership with Nvidia. ML trained on Tempus data BioHive-1 supercom
Slide 33: The result is a palette of ever-evolving sophisticated modules
The result is a palette of ever-evolving sophisticated modules 33
Slide 34: LOWE
) LOWE
Slide 35: The Recursion OS is more than a collection of point solutions
The Recursion OS is now more than a collection of point solutions accessible to expert users …it is increasingly integrated and accessible via a Discovery User Interface that can be used by any of our scientists from the comfort of their laptop… 35
Slide 36: What to Watch for from Recursion: Potential Near-Term Milestones
• Potential for additional INDs • HR-Proficient Cancers RBM39 in H2 2024 • In-licensed program from Bayer (Target Epsilon) for a novel target in fibrotic diseases now entering IND- enabling studies • Expected Ph2 trial starts • Ph2 FPI for AXIN1 or APC mutant cancers program expected in Q1 2024 • Ph
Slide 37: Questions?
Learnings) Call Questions?