Google Cloud puts focus on precision medicine and life sciences with new AI tools

Pfizer and Colossal Biosciences are among the biotech and pharma firms already using two new suites of artificial intelligence tools unveiled this week by Google Cloud.

The new technologies, announced for general availability Tuesday at the Bio-IT World Conference & Expo in Boston, are designed to help organizations advance drug discovery and precision medicine, according to Google.

The Target and Lead Identification Suite is meant to help pharmaceutical researchers better identify the function of amino acids and predict the structure of proteins. The Multiomics Suite, meanwhile, is focused on discovery and interpretation of genomic data, to help with the development and deployment of precision therapeutics.

The Target and Lead Identification Suite offers AI-enabled data ingestion, helping users streamline management and sharing of public datasets and other information with other organizations using Google Cloud’s Analytics Hub. Using Google’s AlphaFold2 and Vertex AI data pipelines, the suite also helps organizations more accurately predict protein structure and boost success rates compared to more traditional methods.

And with more cost-effective high-performance computing resources, the tools help with target discovery, preparation of lead candidates and virtual high-throughput screening – all toward the goal of finding more promising lead candidate molecules for the drug discovery pipeline.

Meanwhile, the new Mutiomics Suite can help customers streamline the ingestion, sharing and management of data, according to Google, with dataset dashboards from Collibra. It also offers secure exchange capabilities with other Analytics Hub users.

The Multiomics tools can help manage raw sequence files with genome-wide association study pipelines into Google Cloud Storage, extract variants using Batch API and NVIDIA’s Parabricks for accelerated genomic analysis with tools like GATK and Google’s DeepVariant, and accelerate processing using Compute Engine, the company says.

In addition to helping turn raw DNA/RNA sequencing data into more actionable and scalable insights, the suite enables Google clients to identify genes associated with particular diseases or disease traits to be integrated into multimodal datasets and helps organizations process and analyze variants.

Google Cloud notes that drug development – from ideation to finished product – can take 12 to 15 years and cost $1 billion, and just identifying the right target for drug intervention can take as long as a year.

Meanwhile, as genomic insights have grown, managing volumes of that data requires expensive and resource-intensive infrastructure strategies for acquiring, storing, distributing and analyzing genomic data as the volume of data doubles every seven to 12 months.

With the Target and Lead Identification Suite and the Multiomics Suite – which were announced this week as AWS unveiled some new omics capabilities of its own – Google says it wants to help its customers manage those two imperatives, said Shweta Maniar, director of industry solutions for Healthcare & Life Sciences at Google Cloud, in a blog post on Tuesday. And she sees advances in artificial intelligence and machine learning as key drivers in those efforts.

That holds true particularly for the fast-evolving applications of generative AI, which Google sees “occupying a uniquely powerful place in the toolset,” said Maniar.

“Generative AI, which moves AI from analysis to creation of new information, is generally based on large models, which look at the billions of interactions that create communications, whether that’s words in speech or pixels in images,” she explained. “Once trained on an LM, generative AI is capable of digesting and summarizing information, and creating new information, all based on sophisticated probabilistic analysis and a lot of specialized computing.”

With so much more to uncover about the “code of life,” she said, “many puzzles lie ahead, and there’s much understanding to be gained. It’s clear that nature won’t be running out of complexity any time soon, so there’s a boundless amount left to learn.”

“At Cerevel we aim to unravel the mysteries of the brain to solve some of the most difficult to treat neuroscience diseases, including evaluating novel therapies for schizophrenia, epilepsy, and Parkinson’s disease,” said Claude Barberis, vice president, Medicinal Chemistry at Cerevel, in a statement. “The Target and Lead Identification Suite with AlphaFold gives us a valuable tool to more efficiently evaluate and improve our discovery efforts.”

“Combining the science of genetics with the business of discovery, we endeavor to advance the economies of biology and healing through genetics,” added Dr. Alexander Titus, vice president of Strategy & Computational Sciences at Colossal Biosciences. “With Google Cloud’s Multiomics Suite, Colossal was able to see 52% reduction in overall cost and an 88% reduction in time to execute whole genome sequence analysis when compared with open-source tools – demonstrating dramatic improvements in overall computational efficiency.”

Mike Miliard is executive editor of Healthcare IT News
Email the writer: [email protected]

Healthcare IT News is a HIMSS publication.

Source: Read Full Article