Frequently Asked Questions
Answers about the PanAum Platform
Direct, factual answers to the questions researchers, biotech founders, and pharma teams ask most about PanAum.
- 01
What is PanAum?
- PanAum is an AI-driven computer-aided drug discovery (CADD) platform that combines multi-omics integration, causal genetics, pharmacogenomics, and generative AI to accelerate the path from disease hypothesis to a population-characterized, IND-ready therapeutic candidate.
- 02
How does PanAum differ from traditional CADD platforms?
- Traditional CADD depends on manual associative methods, virtual library screening, post-hoc ADMET checks, and single-omics data, which typically means 6 to 10 years before reaching IND. PanAum uses causal Mendelian Randomization for target discovery, generative AI for de novo molecule design, in-loop ADMET and pharmacogenomic constraints, and multi-omics integration across more than 10 unified sources, cutting time-to-IND to 18 to 36 months.
- 03
What therapeutic modalities does PanAum support?
- PanAum provides specialized workflows for four modalities: Small Molecules, Antibodies, PROTACs, and Novel Target classes.
- 04
What data sources does PanAum integrate?
- PanAum integrates OpenTargets, GWAS, CTD, PharmGKB, CPIC, gnomAD, and the FDA Table of Pharmacogenomic Biomarkers with Whole Genome Sequencing, Proteomics, and Epigenomics layers, totaling more than 10 unified data sources for deep biological and pharmacogenomic context.
- 05
What are the stages of the PanAum pipeline?
- The pipeline has six automated stages: Evidence Input, Network and Enrichment, Prioritization, Optimization, Generative Design, and Output and Reporting. Full scientific provenance is maintained throughout, and an integrated PGx track runs alongside every stage.
- 06
What is the Generative Fusion Engine?
- The Generative Fusion Engine is a heterogeneous Graph Attention Network (GAT) that synthesizes multi-omics inputs to predict high-affinity molecular scaffolds. It goes beyond traditional screening to enable genuine de novo molecular invention, with pharmacogenomic variant constraints built directly into the generation process.
- 07
How does in-loop ADMET prediction work?
- In-loop ADMET prediction builds safety and pharmacokinetic constraints directly into the molecule generation process, so unsafe candidates are filtered out during design rather than in late-stage testing. This cuts both pre-clinical and clinical failure rates.
- 08
What is the Pharmacogenomics (PGx) Engine?
- The PGx Engine applies population-level variant intelligence from PharmGKB, CPIC, gnomAD, and the FDA Table of Pharmacogenomic Biomarkers at every pipeline stage. Metabolizer phenotype profiling across CYP2D6, CYP2C19, CYP3A4/5, UGT1A1, and SLCO1B1 ensures every lead is safe across real patient populations, not just the reference genome.
- 09
Does PanAum use AlphaFold3?
- Yes. PanAum integrates AlphaFold3 for structure prediction, identifying binding pockets and cavity geometries that feed into generative molecular design. This includes variant conformations arising from common missense polymorphisms.
- 10
Who is PanAum built for?
- PanAum serves three main audiences. Academic researchers can validate disease hypotheses with causal genetic evidence without needing a wet lab or clinical genetics team. Biotech startups can compress discovery timelines and enter investor conversations with a PGx-characterized candidate and Companion Diagnostic strategy. Global pharma teams can integrate AI-driven target intelligence and regulatory-ready PGx reporting into their existing R&D pipelines.
- 11
How long does it take to go from target to IND with PanAum?
- PanAum delivers population-characterized, IND-ready candidates in 18–36 months, compared to 6–10 years for traditional CADD workflows.
Still have questions?
Contact our team for technical documentation and platform access.