• LatchBio provides white-labeled data infrastructure, analysis tools and delivery portals for biotech solution providers.
• In collaboration with Miraomics and Pythia, the company released a 30 million cell atlas spanning 150 diseases, 200 tissues, and 27 technologies available for immediate access on a usage basis.
• The company also released an agentic AI curation framework used to improve efficiency and accuracy of human-in-the-loop molecular data cleaning, with an associated white paper outlining its design and function.

SAN FRANCISCO: Miraomics, Pythia Biosciences and LatchBio released a 30 million cell atlas spanning over 150 indications, 200 tissue types and 27 measurement technologies curated from public sources.

Millions of single cell transcriptomes are scattered across the Internet but remain unused because of the expensive human labor required to structure and annotate this data for downstream use. At this time, these public datasets in aggregate constitute the largest bank of scRNA-seq in existence and the most diverse source of diseases, tissues and patients.

Progress in engineering biology increasingly depends on data-hungry statistical models to reason about emergent properties of living systems that exceed capabilities of unaided human cognition. While purpose-built industrial data generation efforts, like perturbation atlases, offer a path forward, they do not yet sample from enough broad observational data to generalize to many practical translational contexts, especially those addressing indications with small patient populations.

Solution providers, like Pythia and Miraomics, use deep knowledge of molecular data curation to structure publicly available studies for large scale bioinformatics and machine learning. Using Latch’s white-labeled data infrastructure and data portal, they clean and distribute millions of cells through Latch to their biopharma and biotech customers.

“By collaborating with forward-thinking partners like LatchBio and Miraomics, we can bring our high-quality, expertly curated scientific content to a broader segment of the research community and help accelerate life-saving breakthroughs. This marks the first of many releases where portions of the Pythiomics multi-omics database, known for its depth, precision, and scientific rigor, will be conveniently accessible via the Latch platform,” said Tristan Gill, Co-Founder and CEO at Pythia.

“We are excited to announce this major release of high quality curated data, representing thousands of hours of curation effort, enabling new opportunities for development of novel AI tools and novel insights in basic science, disease progression and drug discovery,” said Eugene Bolotin, Co-Founder and CEO at Miraomics.

LatchBio also releases a suite of agentic molecular curation tools that improve per-dataset curation times by around 40x and increase annotation quality and consistency by incorporating information from entire papers and unstructured supplements. This framework can completely automate curation in some cases. A whitepaper detailing its design and function can be found here: http://latch.bio/latch-curate.

“By partnering with leading solution providers, our ambition is to organize the world’s public molecular data for immediate access on a usage basis, for small biotechs, large pharma and frontier AI labs alike,” said Kenny Workman, Co-Founder and CTO at LatchBio.

Fonte: Business Wire