Explore the full de-identified screening results across 34 IMD yeast models and 8,384 TargetMol compounds. Compound identities are anonymized (CPD_xxxxx IDs) until IP is filed; all effect sizes, pathway annotations, and dual-condition co-rescue signals are public.
Rescuer and sensitizer counts for all 34 models, dynamic-range scatter, and pathway hit-rate.
Top 300 compounds by dynamic range. Filter by profile class and minimum |Z|.
Searchable, sortable table over all 8,384 compounds. Compound identities are redacted.
Gene × gene heatmap of shared rescuers and sensitizers — empirical evidence for “platforms-in-a-pill”.
34-gene Z-score fingerprint for any compound, plus the Cure Odysseys posts keyed to its top-hit gene.
This dashboard is running in public mode: all compound names, SMILES strings, and target annotations are redacted. Every row is keyed by an opaque CPD_xxxxx identifier. Perlara team members access the full-fidelity internal version behind SSO.
The portal is rebuilt automatically after every screening batch. Compound-level de-identification is applied at preprocessing — names, SMILES, and explicit targets never reach the public bundle.
The dataset behind this portal is batch TM_03162026: 34 yeast disease avatars × 8,384 TargetMol compounds, processed through Perlara’s standardized growth-rescue assay. Rescuers are scored at Z ≥ 2.5 and sensitizers at Z ≤ −2.5 relative to DMSO control. Each row in the heatmap is one compound; each column is one disease model.The portal is rebuilt automatically after every screening batch. Compound-level de-identification is applied at preprocessing — names, SMILES, and explicit targets never reach the public bundle.