Research account from Biocentric’s published case study. Findings are reported for the methods described; limitations are noted below.
Background
The study investigated TYK2, focusing on the JH2 pseudokinase domain and its relationship to the JH1 kinase domain. Its discovery question was how to identify selective candidate molecules while considering off-target and pharmacokinetic properties.
Research objectives
Identify compounds predicted to bind TYK2 JH2 while avoiding JH1; evaluate interactions with other Janus kinase family proteins; and refine selection using ADMET analysis.
Virtual screening
The source describes screening the ZINC database, approximately 900 million compounds, for candidates with potential to bind the TYK2 JH2 pseudokinase domain.
Selectivity and ADMET analysis
After initial screening, candidates were evaluated against other Janus kinase proteins to investigate potential selectivity. Computational evaluation cannot by itself confirm the absence of off-target effects.
Promising compounds underwent ADMET analysis in collaboration with Toxometris.AI to refine the shortlist around predicted pharmacokinetic and drug-like characteristics.
Reported findings
The original account describes diverse chemical classes and potential scaffolds for binding within the JH2 pocket. It reports that some candidates had greater predicted binding affinity than deucravacitinib, the comparator discussed in the source.
The public case study does not include a full experimental affinity or selectivity dataset. Comparative predictions should be treated as research hypotheses rather than demonstrated therapeutic advantages.
Impact and next steps
The study connects large-scale virtual screening, selectivity assessment and ADMET analysis into one discovery workflow. Future work described in the source includes further optimization and investigation in preclinical studies.
The separate TYK2 workflow diagram references wet-lab testing; the public text for this case does not establish a validated clinical outcome.