Fusão entre Recursion e Exscientia Cria um Novo Líder em Descoberta de Medicamentos com IA

Novos Métodos para Descoberta de Medicamentos

Discovering new drugs has become increasingly expensive and complex in the last few decades, with new therapies costing more than a billion dollars to be developed.

Isso se deve em parte ao fato de que todos os frutos fáceis já foram colhidos, como plantas medicinais conhecidas e bioquímicos “fáceis” de encontrar.

Outro fator é que as doenças ainda sem tratamento eficiente são as mais complexas, frequentemente causadas por disfunções corporais complexas (diabetes, Alzheimer, câncer, obesidade, etc.) ou causas de difícil alcance (por exemplo parasitas e vírus como HIV ou malária, muito bons em evadir o sistema imunológico).

Isso significa que, para descobrir novas moléculas agora, centenas de milhares ou até milhões de compostos precisam ser considerados antes de reduzi-los a poucos.

Esta é uma tarefa assustadora e também um processo caro.

Felizmente, o progresso em IA e computação tornou possível vasculhar um volume astronômico de dados, a um custo muito menor tanto em dólares quanto em tempo.

And two companies are now merging to speed up the adoption of AI-driven drug discovery.

Fusão Exscientia & Recursion

On August 8th, 2024  the merger of Exscientia with its larger peer Recursion Pharmaceuticals was announced.

Both companies had developed their own process to leverage AI and automation to speed up drug discovery, as well as reduce costs.

A strong argument for the merger is that both companies’ technologies are quite complementary. We will look in detail below at both companies, but the overall picture is such:

• Recursion está focada em oportunidades “primeira na doença”, por meio de uma compreensão mais profunda dos mecanismos biológicos.
• Exscientia está focada em medicamentos “best‑in‑class”, com expertise em química de precisão e síntese molecular.

So together, the combined company not only gets some additional scale to save money on overhead costs and regulatory compliance (usually the goal of mergers in biotech) but also the combination of excellence in chemistry synthesis AND biological insights.

Exscientia

The company is using AI to develop terapias de precisão.

It runs a “full stack” AI drug discovery technology with dedicated software at every stage of the drug discovery process.

So instead of looking at existing molecules, Exscientia’s Precision Design AI designs custom molecules to match the target found by its Precision Target AI.

Fonte: Exscientia

Exscientia’s technology reduces 70% of the time required for going from a biological target to finding a corresponding drug and an 80% more capital-efficient process.

Part of the time and cost saving comes from a highly automatized process, with “automação robótica abrangente em todo o ciclo de experimentação”.

This resulted in 4 compounds in early clinical stages, 30 programs in total, and $6.5B in revenues from milestones with partners. The main focus has been oncology (cancer) and inflammatory diseases.

Fonte: Exscientia

Recursion Pharma

Recursion Pharmaceuticals leverages AI in drug discovery. The more AIs get involved in drug discovery and development, the more data will become precious for training the AIs.

Biology is an extremely complex field, with integrated and verified data sometimes in short supply. This is a serious problem when any error will create bias, limitations, and errors in the AI, which might then need to be retrained from scratch.

So creating solid datasets has been the focus of the company since its inception looking to solve several problems with biodata:

• Dados analógicos, de faxes a PDFs ou impressões digitalizadas.
• Dados isolados, com pouca ou nenhuma anotação.
• Pesquisas difíceis de replicar.

To solve these problems, Recursion created one of the world’s largest automated wet labs, and digitized millions of their own experiments (2.2 million experiments per week).

It combines dry lab (in-silico) and wet lab (biological samples) with:

• Uma biblioteca de 1,7 milhão de pequenas moléculas.
• Culturas celulares, edição gênica CRISPR, fatores solúveis, vírus vivos, etc.
• Um fluxo de trabalho de robótica laboratorial automatizada que permite até 2.2 million experiments each week.
• Microscópios de alta capacidade e sistemas de sequenciamento.
• Fluxos de vídeo contínuos de câmeras, registrando medições holísticas de comportamentos animais.
• Recursos computacionais avançados, que geraram >21 petabytes of proprietary high-dimensional data.
• Dados ADMET (absorção, distribuição, metabolismo, excreção e toxicologia).

This creates unique (and massive) datasets at all levels of the “multiomics” biosciences, including proteomics (protein levels), transcriptomics (mRNA levels), phenomics (cellular morphology), ADMET and “in-vivonomics” (animal behaviors). The company is also looking to add metabolomics and genomics to its datasets in the future.

Fonte: Recursion

(você pode ler mais sobre por que multiômicas são importantes em “Multiômicas são o próximo passo na biotecnologia”).

Portanto, enquanto a Exscientia começou estudando mecanismos biológicos e projetando um medicamento para eles, a Recursion, por outro lado, construiu do zero um enorme banco de dados de pesquisas biológicas padronizadas e replicadas.

Recursion também adquiriu em maio de 2023 as startups pré-clínicas focadas em química de medicamentos, Cyclica e Valance, for a total of $87.5M.

They also own one of the world’s fastest supercomputers to train their LLMs and AIs for drug discovery. Models were trained on a library of more than 2 billion images and inferred 6 trillion relationships between all possible combinations of genes and compounds.

Fonte: Recursion

Recursion estabeleceu uma parceria com a líder em IA Nvidia and might release some of its AI models to commercial partners via NVIDIA’s new BioNeMo platform. It will also give Recursion priority access to NVIDIA’s latest GPUs through NVIDIA DGX™ Cloud.

The company has also received um investimento de $50M da NVIDIA em julho de 2023.

Recursion’s R&D proprietary pipeline is mostly focused on rare diseases and oncology, with 3 candidate drugs in phase 2 of clinical trials.

Fonte: Recursion

Para setores mais complexos, como neurociência e oncologia intratável, a empresa prefere estabelecer parcerias com companhias consolidadas nesses setores. Por exemplo, Roche em neurociência e Bayer em alvos de oncologia intratável e fibrose.

No total, a empresa estima $13B in potential milestones across 50+ possible programs plus royalties.

Por fim, a empresa estabeleceu relações para licenciar sua tecnologia e dados, especialmente quando a troca de dados pode ser negociada para ampliar as informações que ambas as empresas podem usar no futuro.

Cenário Pós‑Fusão

Propriedade & Gestão

Overall, while technically a merger, it seems to be somewhat of an acquisition of Exscientia by Recursion.

The merger is organized so that Exscientia shareholders will receive 0.7729 of Recursion Class A common shares. Post-merger, the Recursion shareholders will own 74% of the combined company and Exscientia shareholders 26%.

The transaction should be closed by early 2025.

Recursion’s CEO and co-founder will be the CEO of the combined entity, while Exscientia’s CEO will become the Chief Scientific Officer.

The company location might however become a little bit complex, with >850 employees spread among the many inherited sites in Salt Lake City, London, Toronto, Montreal, San Francisco Bay Area, Oxford, Boston, Vienna, Dundee, and Miami.

Most likely, some relocations and mergers of facilities might be in order in the medium term.

P&D Fundida

The merged R&D pipeline is equally split between oncology and rare diseases, with 4 programs in phase 2 of clinical trials.

Fonte: Recursion

For the future of research at the company, many of the research steps can be supplemented by the merger of recursion and Exscientia data and AI tools, especially in the initial stages.

For example, Exscientia’s automated chemistry synthesis mix is a strong addition to Recursion’s world largest automated wet lab.

The deep understanding of biology and disease of Recursion can also be used to better pick what is useful in the millions of physical compounds and billions of target predictions of Exscientia.

Fonte: Recursion

In addition to this proprietary pipeline, the company will also have 4 large strategic collaborations (e.g., Roche, Bayer, Sanofi, Merck KGaA) with 10 programs already optioned across oncology and immunology.

It is likely that the wider range of ongoing collaboration will also give the company’s management some leverage in future strategic collaboration discussions, as they will now have additional options to put the large pharmaceutical companies in competition with each other.

Fonte: Recursion

Dados Financeiros

Both companies are mostly pre-revenues, with no drug on the market yet. The existing cash comes from successive fundraising and income mostly from reaching milestones in research programs established with larger pharmaceutical companies.

The combined business will have a $850M treasure chest of cash, giving a runway until 2027.

The next 2 years should see up to $200M in potential milestone payments for R&D successes. Most of the research programs are targeting large markets, with the majority of them aiming for >$1B in sales if successful.

O Futuro da Descoberta de Medicamentos

It is a growing trend that AI and big data will be key in finding new cures and better drugs in the future.

It is also becoming very clear from LLMs and other AI initiatives that data is king in this sector.

So we should be expecting mergers like the one of Exscientia & Recursion to keep happening in the future. We discuss many of these in our articles “Top 10 Empresas de Biotech Big Data” and “Top 5 Empresas de IA & Biotech Digital”.

Among the other dataset that could be useful in the future to Recursion, or other similar AI-drug discovery companies are:

• Genômica espacial, ou o estudo do genoma e transcriptoma em 3D, permite a visualização da atividade dos genes ao nível celular ou até intracelular.
  • 10x Genomics é agora a principal empresa independente do setor desde que seu concorrente Nanostring foi adquirido pela Bruker Corporation.
• Modelagem molecular baseada em física.
  • Schrödinger (SDGR) é líder em encontrar a melhor molécula possível para um determinado objetivo, equilibrando métricas conflitantes como potência, solubilidade, meia‑vida, sintetizabilidade, etc.
• Leitura e sequenciamento de DNA
  • As empresas líderes são Illumina (ILMN), Pacific Biosciences(PACB), e Oxford Nanopore (ONT.L)
• Biologia sintética, ou a engenharia sob demanda de novas características ou organismos para propósitos específicos, seja médico, biotecnológico ou industrial.
  • A empresa líder no setor é Ginkgo Bioworks(DNA).