Pythia Biosciences and Miraomics have announced a strategic collaboration to develop Pythiomics—a rich, high-quality and harmonized single-cell omics database designed to accelerate discoveries and AI/ML development in drug research. Pythiomics combines expert manual curation, rigorous quality control, metadata harmonization, and cutting-edge AI techniques for cell type annotation. The result is a standardized, reliable, and accessible data resource that enables biopharmaceutical companies and research institutions to validate insights, develop AI/ML models, and accelerate target and biomarker discoveries.

Despite the explosion of single-cell datasets in recent years, much of the data are stored in inconsistent formats and annotated with non-standardized metadata, while scattering across different repositories. These issues make curation, integration and analysis time-consuming and costly—often placing the power of single-cell data out of reach for many research efforts.

To solve this, Pythia Biosciences and Miraomics have partnered to build Pythiomics with a focus on scientific rigor and usability. The database follows a step-by-step standard operating process (SOP) to ensure rigorous quality control, metadata harmonization and comprehensive documentation for traceability. AI-powered tools are employed to assist in accurate cell type annotation and metadata harmonization, while all data is mapped to widely adopted public and proprietary ontologies.

As of May 2025, Pythiomics comprises 107 million cells spanning hundreds of disease subtypes and incorporating more than 20 standardized vocabularies. The database is available through the C-DIAM Multi-Omics Studio, a web-based platform developed by Pythia Biosciences that supports interactive exploration of multi-omics datasets—including single-cell, CITE-seq, Visium, Visium HD, Xenium, bulk RNA-seq, proteomics, and metabolomics data. Pythiomics is also available via AWS S3 delivery.

“The challenges of harmonisation, annotation and accessibility of multi-omic data create a major bottleneck in addressing critical scientific questions. Our collaboration is a pivotal step towards addressing this issue and unlocking this critical resource, enabling researchers to accelerate drug discovery and innovation,” remarked Dr. Eugene Bolotin, CEO of Miraomics.

“Miraomics brings unparalleled expertise in leveraging both in-house and vendor data to drive drug discovery and development in Pharma. Recognizing that data quality is paramount and understanding where other data sources have fallen short, they have played a pivotal role in establishing optimal standard operating procedures to ensure we generate the highest-quality data possible.” said Tristan Gill, CEO of Pythia Biosciences. “By combining Pythia and Miraomic’s deep knowledge of data curation with advanced AI tools, we are making significant strides in harmonizing complex datasets and empowering the research community.”

In addition to single-cell transcriptomics data, the Pythiomics database also incorporates multi-omics data, including proteomics, bulk RNA-seq, and metabolomics – totaling more than 10,000 datasets. This integration enables scientists to cross-validate insights across omics layers, offering a holistic perspective on disease mechanisms and therapeutic opportunities.

To learn more about the collaboration and upcoming disease areas to be added to the Pythiomics database, visit www.pythiabio.com