In the last year, the buzz around artificial intelligence/machine learning (AI/ML) tools has grown to a deafening roar, with applications in new and exciting industries. In the life sciences sector – where more ‘omics data is generated than can be effectively analyzed – building a scalable, reproducible workflow remains a serious priority, yet an extraordinary challenge. Layering in any sort of AI element to this process, increases the complexity of these processes significantly. For startups to mid-sized and large companies alike, creating an infrastructure and skilled workforce for complex multi-omics analysis traditionally has taken time and considerable capital investment to build the right team.
Enter Google Cloud's Multiomics Suite, making it easier for life science teams to harness the power of advanced bioinformatics and AI for integrated genomic, transcriptomic, proteomics, and other data analysis. For those in the drug discovery and precision medicine space, these tools are game changers. "These new solutions launching today can transform life sciences organizations by accelerating drug discovery and bringing therapeutics to market faster,” says Shweta Maniar, Director of Industry Solutions for Healthcare & Life Sciences at Google Cloud in a press release announcing the Suite’s launch. “When patients are waiting for that life-saving treatment in cancer care or that quality-of-life medicine for migraine headaches, this faster time-to-market can have an incredibly positive impact on lives."
Overall, the platform's ease of use offers increased access to advanced computational techniques so that discovery teams can get more from their data and increase research efficiency. In collaboration with Form Bio, as a Google Cloud Premier partner, researchers can bring their solutions, including cell and gene therapies and other biologics, to market faster than ever before, giving patients a chance at a brighter future.
Multiomics Discovery, Data Sharing, and Analysis Enters a New Age of Precision Medicine
Multiomics data analysis requires the integration of distinct data sets from genomics, transcriptomics, proteomics, and other techniques to create more effective targeted therapeutics. While complicated and time-consuming to execute, the pay-off is huge: A comprehensive understanding of the cellular link between genotype and phenotype can rapidly accelerate drug discovery efforts or biomarker selection for patient stratification.
Due to the above mentioned challenges, most multiomics data is not adequately mined for novel findings. Put simply, it’s (relatively) easier for researchers to generate genomic, transcriptomic, proteomic, or epigenomic data than to spend the time building a bioinformatics pipeline or training and validating an AI algorithm to extract actionable insights from this tangled mess of complex data.
Current AI/ML algorithms demand massive compute power and electricity, exacerbating existing problems. OpenAI reports that AI training compute usage has doubled every 3.4 months since 2012. Unfortunately, inadequate computational power dampens analytical effectiveness and prolongs bioinformatics pipeline or AI workflow run times. On top of that, managing the various components involved in developing, training, evaluating, and serving any AI/ML algorithm (MLOps) adds significant complexity, making adoption a daunting task for life science teams. But fortunately, by employing Google’s Multiomics Suite that leverages Form Bio’s expertise and workflows, such teams can mitigate the issues, reduce the complications caused by legacy platforms, enhance research pace, and gain scientific insights faster.
Google Cloud's Multiomics Suite provides access to significant cloud-enabled computational power, empowering users to accelerate their multiomics analyses and alleviating the pressure of having to implement AI/ML system operations. Compared to the use of open-source bioinformatics tools, our parent company, Colossal Biosciences, reduced whole-genome sequencing analysis time by 88% and cost by 52%, dramatically improving the overall computational efficiency. Moreover, due to Google’s leading role in the industry as the cleanest cloud service, running bioinformatic workflows and applications Google Cloud’s Multiomics Suite, requires zero net emissions, 100% renewable energy, and minimal energy consumption for machine learning. Combining reduced costs, increased efficiency, and environmental-friendly computing power, Form Bio and Google are pushing forward the development of the multiomics analysis field, elevating it to new heights.
Data In, Insights Out
In addition to providing access to AI workflows, the Multiomics Suite also serves as a central data sharing and management repository. It’s cloud agnostic, allowing organizations to choose their own cloud solution, scale at their own pace, and incorporate open-source or proprietary tools for custom analyses.
AI models from the Suite are fully traceable and reproducible, powered by Vertex AI. This enables users to organize millions of artifacts in their cloud environments. Furthermore, it leverages the power of Google Research to transform hundreds of thousands of files, samples, and records to load variant call format (VCF) files from Google Cloud Storage into BigQuery for analysis.
Making Streamlined AI a Reality in the Gene Therapy Sector
The rapidly expanding field of AI/ML tools has captured significant attention across various industries, including the life sciences sector. Google Cloud’s selection of Form Bio and other partners to help with the Multiomics Suite, will streamline the adoption of AI and bioinformatics tools in the cell and gene therapies space, advancing powerful therapeutics for those living with rare diseases. AI has enabled target selection for CRISPR-based therapeutics, optimization of viral vectors, and more streamlined pre-clinical and clinical testing. The ultimate benefit of this is more rapid development of therapeutics that can be tailored to individual patients.
At Form Bio, our commitment to the streamlined and accessible nature of AI has already started with our new product, FormSightAI. This AI-driven platform reduces time and cuts costs associated with the biomanufacturing of AAV vectors, used to deliver powerful gene therapies. By using FormSightAI for candidate validation throughout preclinical research, drug developers can proactively predict and address manufacturing issues, before they rear their ugly heads the late stages of clinical testing.
AI Disclosure: Feature image was generated by AI image tool MidJourney.