Old Drugs, New Targets: Multi-parameter computational prioritization of FDA‑approved, brain‑penetrant drugs as isoform‑selective GSK‑3β inhibitors for Alzheimer’s disease
Abstract
Glycogen synthase kinase-3β (GSK-3β), a central regulator of tau phosphorylation and amyloid pathology, with its β isoform (GSK-3β) used as a main driver of the Alzheimer’s disease (AD) neurodegeneration. It is a validated therapeutic target in AD, yet no isoform-selective, brain penetrant inhibitors have reached clinical use. This paper screens 1000+ FDA-approved drugs by dual-isoform docking (GSK3α/β) and 8-seed redocking, prioritizing compounds by its α/β selectivity (G_select = G_GSK3A − G_GSK3B), CNS penetration, Hinge H-bonds, and pose plausibility. High selectivity alone provided misleading: filgotinib showed the largest G_select (+1.25 ± 0.03 kcal/mol), yet a reproducibly implausible pose, while abemaciclib (ΔG_select = +0.80 ± 0.09 kcal/mol; plausible‑pose fraction = 0.88) and peficitinib (+0.008 ± 0.02 kcal/mol; 1.00) provided a moderate selectivity with structurally reliable binding. Multi-seed redocking led to a hidden instability (loratinib plausible in only 25% of the seeds). This work provides a ranked, hypothesis-generating shortlist by establishing that isoform-selective GSK3β inhibitor discovery demands joint optimization of selectivity, and multi-seed pose stability, rather than affinity of GSK-3β alone. The following results are computationally prioritized candidates, rather than experimentally verified GSK-3β inhibitors.
Related
- Benchmarking Affinity-Based Docking for Irreversible Acetylcholinesterase Inhibitors: A Physics-Informed Reactivity Framework for Stereoselective Potency Prioritization
- Computational design and prioritization of database-absent small-molecule GLP-1 receptor agonist candidates from public bioactivity data
- Ligand-Based Design of Lyn-Selective Type II Kinase Inhibitors by Rational Tail Modification of Imatinib
- Discovery of selective, metabolically stable pyrazole-based FLT3 inhibitors for the treatment of acute myeloid leukemia
- A Computational Workflow Integrating Explainable AI and Database Mining for the Prioritization of Antimicrobial Natural Products
BibTeX
@misc{baljinnyam2026drugs,
title = {{Old Drugs, New Targets: Multi-parameter computational prioritization of FDA‑approved, brain‑penetrant drugs as isoform‑selective GSK‑3β inhibitors for Alzheimer’s disease}},
author = {Naranbileg Baljinnyam and Ariuntuguldur Khurelbaatar},
howpublished = {ChemRxiv},
year = {2026},
month = oct,
doi = {10.26434/chemrxiv.15010468/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010468/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Old Drugs, New Targets: Multi-parameter computational prioritization of FDA‑approved, brain‑penetrant drugs as isoform‑selective GSK‑3β inhibitors for Alzheimer’s disease
AU - Naranbileg Baljinnyam
AU - Ariuntuguldur Khurelbaatar
PY - 2026
DA - 2026/10/14
PB - ChemRxiv
DO - 10.26434/chemrxiv.15010468/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15010468/v1
AB - Glycogen synthase kinase-3β (GSK-3β), a central regulator of tau phosphorylation and amyloid pathology, with its β isoform (GSK-3β) used as a main driver of the Alzheimer’s disease (AD) neurodegeneration. It is a validated therapeutic target in AD, yet no isoform-selective, brain penetrant inhibitors have reached clinical use. This paper screens 1000+ FDA-approved drugs by dual-isoform docking (GSK3α/β) and 8-seed redocking, prioritizing compounds by its α/β selectivity (G_select = G_GSK3A − G_GSK3B), CNS penetration, Hinge H-bonds, and pose plausibility. High selectivity alone provided misleading: filgotinib showed the largest G_select (+1.25 ± 0.03 kcal/mol), yet a reproducibly implausible pose, while abemaciclib (ΔG_select = +0.80 ± 0.09 kcal/mol; plausible‑pose fraction = 0.88) and peficitinib (+0.008 ± 0.02 kcal/mol; 1.00) provided a moderate selectivity with structurally reliable binding. Multi-seed redocking led to a hidden instability (loratinib plausible in only 25% of the seeds). This work provides a ranked, hypothesis-generating shortlist by establishing that isoform-selective GSK3β inhibitor discovery demands joint optimization of selectivity, and multi-seed pose stability, rather than affinity of GSK-3β alone. The following results are computationally prioritized candidates, rather than experimentally verified GSK-3β inhibitors.
ER -