PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
February 21, 2026Biophysical Journal0 citations

BPS2026 – AI-driven virtual drug discovery of multi-target compounds against breast tumors and Alzheimer’s

View Full Paper
SPSophie J. PrestonAMAdeline MaiJMJuliana Castro Marrero

Key Points

  • To discover multi-target compounds that can inhibit targets related to breast cancer and Alzheimer’s disease.
  • Used AI-assisted virtual drug screening to generate chemical analogs.
  • Performed molecular docking using AutoDock Vina on three targets.
  • Employed SCORCH for rescoring and dynamic thresholding to identify top binders.
  • Identified four potential multi-target compounds including exemestane and its analogs.
  • Results suggest efficacy against ER and tau aggregation.
  • Dynamically evaluated binding affinity and disaggregation behaviors planned.

Abstract

Selective estrogen receptor modulators (SERMs) and aromatase inhibitors (ARIs) are currently used for treating estrogen receptor (ER) positive breast cancers. Interestingly, some of those SERM and ARI drugs have beneficial effects in preventing Alzheimer's via their estrogenic effects in the brain. However, the impact of those anti-breast cancer drugs on early tau aggregation in Alzheimer's is not –studied. Using AI-assisted virtual drug screening, we aim at discovering multi-target compounds (MTCs) that can simultaneously bind to three molecular targets associated with breast cancers and early tauopathies. These targets are ER-α, GSK-3β, and tau-k18 dimer. Here, ER-α is a major subtype of ER, the kinase GSK-3β is responsible for creating misfolded tau, and the tau-k18 dimer is a model of early tau aggregates. Our first step involves generating ∼800 chemical analogs of experimentally derived inhibitors of ERα, GSK-3β, and tau aggregation using chemical similarity search tools. We then successfully performed high-throughput molecular docking of those chemical analogs and their parent compounds against our three molecular targets via AutoDock Vina. Our molecular docking results comprised 20 docked conformational poses for each compound-target complex. Finally, we employed a recently developed supervised machine-learning rescoring tool, SCORCH, followed by the application of a dynamic threshold algorithm, to identify four MTCs that are top binders to our three protein targets. Our preliminary screening results indicate that exemestane, an exemestane analog, a raloxifene analog, and an apomorphine analog are potential MTC candidates for inhibiting ER and tau aggregation. Following these results, all-atom molecular dynamics simulations under physiological conditions will be performed to evaluate the free-energy binding affinity and protein disaggregation behaviors of the newly discovered MTCs in tau aggregates.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Preston et al. (2026) studied this question.

synapsesocial.com/papers/69990e0a5b97ab4c14ac2fddhttps://doi.org/10.1016/j.bpj.2025.11.2606
Ask AI
Helpful
Bookmark
Share
View Full Paper