AIBERT is Precision Scientific’s proprietary multi-omics data analytics platform. AIBERT is designed to bridge the ever-increasing gap between big data and decision-making information for drug R&D.
With big data resource and cutting-edge AI algorithms, the platform focuses on realizing the promise of private data from biopharma/biotech companies, most of which have following three characteristics:
Until recently, we have collaborated with more than 40 biopharma and biotech companies worldwide, mainly in biomarker discovery, indication selection & expansion, and target prioritization based on omics data from pre-clinical or early clinical studies.
AIBERT has collected PB-level omics data of cancer genomics, functional genomics, and pharmacogenomics from cancer patients, PDX models, and cell lines. All data have been well curated, cleaned, and normalized. For cancer genomics data, it covers more than 50,000 patients with approximately 70 cancer types. Genome-wide genetic perturbation data from about 2,000 cancer cell lines are also included.
AIBERT will generate comprehensive and even advanced molecular features at gene level, pathway level, omics level, as well as tumor microenvironment level, based on the private omics data from biopharma and biotech companies. The system will robustly identify the features related to drug response as many as possible through the statistical techniques optimized for small sample size scenarios and taking diverse clinical confounding factors into consideration.
AIBERT will assess all the signals detected from private data one by one based on multi-dimensional datasets and different algorithms of distinct rationales. The most consistent signal will be considered as the most promising one and will be recommended to the biopharma and biotech companies for drug R&D decision-making.
The core team members of AIBERT are from world-class institutes, including MD Anderson Cancer Center, John Hopkins University, Princeton University, Chinese Academy of Sciences, Peking University and so forth. Meanwhile these members have very diverse backgrounds from artificial intelligence, statistics, bioinformatics, population genetics, pharmacology and molecular biology. Our world-class talents with diverse backgrounds are driven by a single passion for unprecedentedly accelerating drug R&D process via big data plus AI algorithms.
The core value of the omics data analytics platform is to maximize the value of your data through driven by deluge of omics data and cutting-edge AI algorithms.
Functional genomics CRISPR/Cas9
Based on customer's needs, design targeted programmes and provide individually adaptable analysis reports.
Pharma P has Drug D, which has just finished Phase I trial. This is a promising pipeline but in a crowded track. Besides several domestic competitors, an exactly similar drug from a MNC is expected to be approved soon. So Pharma P needs a differential competition strategy for commercializing drug D, including biomarker identification, indication expansion and drug combination.
Step 1. Harmonized all the molecular and clinical data on Drug D provided by Pharma P and employing optimized statistical techniques for small sample size and Bayesian inference to detect biomarker candidates and discover potential MoA of Drug D.
Step 2. Evaluated all positive results obtained at Step 1 based on multiple datasets (knowledge graph, genomics data of cell lines, mouse models and patients) and diverse algorithms (rationales from system biology, statistics learning or deep learning). The features with consistent signals will be considered as the most reliable outcome.
Step 3. Explored promising indication expansion and combination strategy by explainable artificial intelligence algorithms which integrated the information from Step 2. Tens of thousands of compound molecules and multi-omics data were collected to train the related algorithms to ensure the accuracy and robustness of predictions.