Pfizer site in Cambridge, Massachusetts / Courtesy of Pfizer

Artificial intelligence and machine learning (AI/ML) are essential to enable drug discovery and development, and Pfizer leads the biopharmaceutical industry into the next wave of innovation. The company is growing rapidly and recruiting talent for a collaborative effort to deliver transformative medicines to patients faster.

The mandate is “uncompromising, very high-quality science,” said Sandeep Menon, Scientific Director, AI Digital Sciences, SVP and Head of Early Clinical Development. BioSpace.

The vision is threefold: to discover the biology of diseases with AI; use this knowledge to design the right molecules; determine the right patient population for successful clinical trials.

“We are building the next generation of tools for use across the spectrum of preclinical and clinical development,” said Jared Christensen, vice president and head of early clinical development, clinical AI/ML and quantitative sciences.

Pfizer is building an “ML Research Hub” to create new predictive models and tools in what it called “a key investment.”

This team, led by Enoch Huang, VP, Machine Learning and Computational Science, will partner with industry experts to ensure successful application of AI/ML by designing, deploying and maintaining tools and techniques. peak. It will uncover information related to the pathophysiology of the disease and generate relevant and testable hypotheses. ML Research’s efforts will be led by Djork-Arné Clevert, who recently joined the company.

“AI/ML was sewn into the fabric of drug discovery at Pfizer,” Huang said. “A sign of success is when our project teams or our design chemists studying compounds use machine learning without knowing that they are using machine learning. This is what happens behind the scenes. »

“However, we need to apply AI/ML beyond drug design, starting with the patient in mind,” Huang continued. “We see tremendous potential for leveraging public and proprietary datasets using ML methods to better understand disease pathophysiology, which could potentially lead to breakthrough efficacy for patients who significantly change their life.”

Many therapeutic applications

The innovation from the collaboration will be therapeutically agnostic, Christensen shared.

“We will start in areas where we are already established,” he said. Pfizer’s main therapeutic areas are internal medicine, inflammation and immunology, oncology, vaccines and rare diseases.

The elevator will be lighter in oncology, where advances in precision medicine have already been considerable. Pfizer plans to build on these gains to better understand patient populations and stratifications, Christensen noted.

“We are looking for data-rich insights to train the models. The opportunity before us is to inform and influence target prioritization and patient stratification with AI/ML, much like we did it in chemistry,” he said.

In internal medicine, Christensen highlighted heart failure, diabetes, and nonalcoholic steatohepatitis where there are large populations and more data is accumulating every day. The same goes for inflammatory and immunological diseases such as rheumatoid arthritis, Crohn’s disease and ulcerative colitis.

Pfizer intends to leverage this data, along with relevant biomarker and next-generation sequencing datasets, to better understand where its drugs can have the most impact.

“I strongly believe that the diseases we now call one thing will continue to subdivide based on biomarkers and clinical phenotypes,” Christensen said. “I believe this kind of revolution will continue to happen in other diseases, similar to what we have seen in oncology. We are trying to catch and ride this wave.

It might not be too long before the waves crest either.

These and other clinical use cases will help guide the development of the methodology within the ML Research Hub. Subha Madhavan was recently recruited as the Head of Clinical AI/ML and Data Science within Early Clinical Development to help define baseline requirements for drug programs that will leverage the Hub’s innovative methods to accelerate the development.

These efforts will use historical clinical trial data, biomarker data, and real-world evidence such as electronic medical records to precisely define patient populations to inform study design.

This is ultimately about improving the likelihood of technical and regulatory success of Pfizer’s clinical trials, Madhavan said.

“Within clinical AI/ML, we are really driving a paradigm shift in precision medicine. We focus on using multimodal data to inform trial design, first-in-human studies [and] our sign of clinical activity studies.

Pfizer applies advanced methods such as classical and deep learning to molecular datasets gathered from its own clinical trials and published studies “to identify patient subpopulations that may respond better to a certain treatment. “, she explained.

“I’m very optimistic about our ability to exploit large, multi-modal datasets and rapidly develop algorithms to predict a variety of patient outcomes.” She predicts that the impact of many of these innovative new tools will reach patients within the next three to five years.

Madhavan was drawn to Pfizer by the company’s “quick thinking” and “cross-functional” approach to drug development.

“Pfizer is a company that, despite being a ‘big pharma’ company, can pivot very quickly, as demonstrated by COVID vaccines and antiviral programs in response to the global pandemic,” she said. “The culture has changed into a culture where we can enjoy these [cross-functional] teams and bring innovation across multiple therapeutic areas.

“We take a disciplined product development approach to define the business value, key stakeholders, core functionality, and usage of each AI/ML model to align with and accelerate our portfolio,” said she added.

State-of-the-art AI/ML application

Pfizer is also applying AI/ML to digital medicine in the Pfizer Innovation Research (PfIRe Lab). Here, researchers are developing algorithms for wearable devices to help scientists monitor symptoms, assess health, and better understand how treatments work.

Wearable devices provide researchers and physicians with “a complete and continuous picture of the patient’s experience” during the evaluation period, Menon said, rather than relying on the patient’s memory during a single visit to the firm.

Cutting-edge advances abound in AI/ML, but Pfizer is particularly interested in those that can help it reach patients with cutting-edge medicines. As Menon said, “We don’t use AI as just a fancy term or a shiny object. It’s about tangible, executable solutions to key research questions.

Christensen highlighted explainable AI as an area that can help build science around disease.

“We’re looking for new computational models that are less black box and more open to understanding what’s going on under the hood,” he said.

When it comes to understanding the molecular basis of disease pathophysiology, Huang pointed to a powerful ML architecture called Transformer, which was developed with language models in mind. Transformer is the basis of Google Translate.

“It can help us understand biomedical literature through natural language processing, which is an area that we are very interested in at Pfizer,” he said.

Madhavan said knowledge graphs can help connect genes to diseases and drugs and help identify new biomarkers associated with certain disease pathways. Knowledge graphs can also draw connections between patient phenotypes and allow researchers to develop more effective treatments for these patient groups.

Bilingual Data Scientists Wanted

As Pfizer bolsters its AI/ML-focused teams, the company is “recruiting” data scientists.

A dynamic mission requires a dynamic mindset that can sort through complex data to make the right scientific and therapeutic connections.

“There are a lot of data analysts out there who have amazing skills, but we need to marry those skills with people who understand the science and are willing to take a scientific lens to those assumptions,” Christensen said. And with a strategic vision that spans so many divisions and specialties, “a communicative and collaborative mindset” is another key attribute.

Madhavan noted that Pfizer is looking for bilingual data scientists with a deep understanding of both data science and clinical science. The successful candidate will have “work experience where they have effectively deployed their quantitative skills to answer key clinical and/or biological questions,” she said.

In the ML Research Center, Huang is also looking for ambidextrous scientists who not only have expertise in ML research and data engineering, but also understand chemistry and molecular/cellular biology.

As all scientists know, the key to any experiment is reproducibility. “If that doesn’t happen in the real world, I think you’re going to do more harm than good,” Menon said.

He stressed the importance of responsible AI. “Often AI is used as a buzzword. It’s basic statistical modeling and mathematical modeling, but if it’s not done by subject matter experts who are experts in science, it will be a weapon that will turn against us.

Christensen presented the opportunity to potential team members: “It’s early days, but it’s a great opportunity for data scientists who want to build foundational and enduring systems to help us make data-driven decisions. the data to innovate across the drug development paradigm.”

Those interested in joining the Pfizer team can find more information here.