ASK

Standigm ASK™ is an AI-aided drug target identification platform developed by Standigm, designed to help pharmaceutical and biotech researchers identify novel, customized drug targets efficiently and effectively. The platform is built for research teams seeking to move beyond labor-intensive manual data collection and guesswork, offering a comprehensive, evidence-backed approach to target prioritization for any disease or phenotype of interest.

At its core, Standigm ASK™ combines a richly curated biological knowledge database, multiple harmonized AI algorithms, and an interactive user interface to deliver not only a prioritized list of target candidates but also the supporting evidence and contextual information surrounding each target-disease relationship. The platform is fully customizable to each customer's research context, making it suitable for a wide range of drug discovery programs.

Database: Standigm BioMap

• Covers approximately 30,000 targets encoded in the human genome
• Includes around 10,000 phenotypes — diseases and abnormalities defined in the Experimental Factor Ontology
• Contains approximately 5,000 approved drugs and drugs from clinical trials
• Holds over 4.3 million total relations across 25 types of known relationships between 8 types of nodes
• Integrates data from 36 public curated source databases
• Incorporates insights from 97,000 scientific publications containing gene-disease associations
• Publication-derived information is extracted using Standigm's proprietary natural language processing (NLP) technologies
• The knowledge graph contains more than 4 million edges and 70,000 nodes, with three main node types: diseases, targets, and compounds

Prioritization Algorithms

• A knowledge graph-based deep learning model for target ranking
• Natural language processing technologies to extract and interpret literature-derived relationships
• A genotype-phenotype predictor based on genome-scale metabolic models
• A random-walk model based on a modified protein-protein interaction (PPI) network and omics data
• Multiple AI approaches are harmonized according to the specific research context of each project
• Causal relationships between diseases and genes are classified as simple associations or direct causations, using experimentally proven data from genetic modification studies

Customization Capabilities

• Database modification: customers can insert or remove information and integrate context-specific data into the existing knowledge graph
• Customer-provided input: AI models can be trained starting from unbiased, customer-supplied data
• Customer-suggested modules: new analytical modules can be constructed in collaboration with the customer
• In-house databases can be built and integrated with public databases as training data for AI models
• Various filtering templates are available to define the desired level of target novelty

Elaboration and Analysis Modules

• NLP technology: identifies co-occurring patterns of specific word types and investigates information regarding particular disease-target pairs
• Mechanism of Action (MoA) investigation: identifies enriched pathways for designated gene sets, such as downstream affected genes for a selected target
• Path analysis: identifies important paths connecting specified disease-target pairs and paths containing one or more designated genes
• Expression analysis: identifies expression patterns of designated gene sets in disease-specific data, including patient samples or cell lines

Interactive User Interface

• Served with a fully interactive user interface for intuitive exploration
• User-driven exploration of prioritization evidence and reasoning
• Command-free database browsing for accessible, efficient research

Workflow and Turnaround

1. Target candidates are selected using AI models applied to the integrated knowledge graph and customer-provided data
2. A filtering step is applied to refine candidates to the desired level of novelty
3. A final target candidate is typically delivered within two weeks

Standigm ASK™ is available as an AI SaaS and API service, and Standigm also offers partnership and collaboration arrangements. The platform supports trial access upon request, and prospective users are encouraged to contact Standigm directly to explore how the platform can be tailored to their specific drug discovery needs.