MILPITAS, Calif. — Sage-N Research, Inc., a deep tech and services company, introduces the world’s first professional AI-enabled deep data platform for clinical proteomics. Proteomics can save significant time and money in clinical trials and discovery by probing protein pathways, but current software — mostly academic PC downloads — lack useful sensitivity and are prone to costly false positives. AIMS™ (“AI for MS”) uses intelligent data mining to increase sensitivity by orders of magnitude while eliminating false positives, which unlocks mass spectrometry’s (MS) potential as the digital extension of the nearly $20B immunoassay (ELISA) market.
“Just as a 3D CT scan reveals more subtle features by combining multiple 2D x-rays, AIMS triangulates many relevant data files to characterize ultra-low abundance proteins,” according to David Chiang, Sage-N Research founder and chairman. “Our breakthrough technique was recently outlined in well-received scientific posters.”
Biopharmaceutical R&D experiments are costly with low productivity because researchers cannot see what is happening in protein pathways when things go wrong. Immunoassays can chemically probe specific protein forms, but they are generally unavailable for low-abundance proteins. In theory, proteomics is a game-changer that can probe practically any MS-detectable protein form, as a representative peptide, using only its amino acid sequence plus post-translational modifications (PTMs). In other words, it is essentially a universal, digitalized immunoassay (“cyber-assay”). In practice, it has been beset by experimental software using non-rigorous statistics that yield often-irreproducible results.
High-value clinical projects can be accelerated and saved with professional-quality, high-sensitivity proteomics. Current software analyze only fragment data (DIA or DDA) from only the 2% most abundant peptides. AIMS uses these identified peptides to mine the 98% scarce peptides for possible variants exactly one PTM mass apart. Since each MS run yields content-rich data both voluminous yet incomplete, sensitivity increases exponentially with each additional data file. In other words, high-sensitivity analysis means big data mining. Because low-abundance protein forms have very few data-points, they are straightforward to analyze once isolated, but isolating them requires professional tech systems like AIMS.