De Novo Design of SIK3 Inhibitors via Feedback-Driven Fine-Tuning of Seq2Seq-VAE

Alzheimers disease (AD), a progressive neuro-degenerative disorder, currently lacks effective therapeutic strategies that can modify disease progression. Recent studies have highlighted the circadian rhythm critical role in AD pathophysiology, implicating circadian clock kinases, such as the Salt-Inducible Kinase 3 (SIK3), as promising therapeutic target. Generative AI models have surpassed traditional methods of drug discovery, untapping the vast unexplored chemical space of drug-like molecules. We present a sequence-to-sequence Variational Autoencoder (Seq2Seq-VAE) model guided by an Active Learning (AL) approach to optimize molecular generation. Our pipeline iteratively guided a pre-trained Seq2Seq-VAE model towards the pharmacological landscape relevant to SIK3 using a two-step framework, an inner loop that iteratively improves physiochemical properties profile, drug likeliness and synthesizability, followed by an outer loop that steer the latent space towards high-affinity ligands for SIK3. Our approach introduces feedback-driven optimization without requiring large labeled datasets, making it particularly suited for early-stage drug discovery in under-explored therapeutic targets. Our results demonstrate the models convergence toward SIK3-specific small molecules with desired properties and high binding affinity. This work highlights the use of generative AI combined with AL for rational drug discovery that can be extended to other protein targets with minimal modifications, offering a scalable solution to the molecular design bottleneck in drug design.

• Publication year

  2025

• Venue

  Unknown venue

• Publication date

  2025-11-10

• Fields of study

  Biology, Medicine, Computer Science

• Identifiers
  arXiv 2511.06930
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Evolutionary and geometric signatures reveal ligand-binding sites across proteomes

  2025cited by this paper
• Optimizing drug design by merging generative AI with a physics-based active learning framework

  2025cited by this paper
• The structures of salt-inducible kinase 3 in complex with inhibitors reveal determinants for binding and selectivity

  2024influential reference
• PubChem 2025 update.

  2024cited by this paper
• Recent advances in Alzheimer’s disease: mechanisms, clinical trials and new drug development strategies

  2024influential reference
• An AI-Driven Framework for Discovery of BACE1 Inhibitors for Alzheimer’s Disease

  2024cited by this paper
• Clinical trials of new drugs for Alzheimer disease: a 2020–2023 update

  2023cited by this paper
• Targeting SIK3 to modulate hippocampal synaptic plasticity and cognitive function by regulating the transcription of HDAC4 in a mouse model of Alzheimer’s disease

  2023cited by this paper
• AI’s potential to accelerate drug discovery needs a reality check

  2023cited by this paper
• AmberTools

  2023cited by this paper
• Alzheimer's disease drug development pipeline: 2023

  2023cited by this paper
• Circadian dysfunction and Alzheimer's disease – An updated review

  2022cited by this paper
• Structure-based drug design with equivariant diffusion models

  2022cited by this paper
• Circadian clocks, cognition, and Alzheimer’s disease: synaptic mechanisms, signaling effectors, and chronotherapeutics

  2022cited by this paper
• Highly accurate protein structure prediction with AlphaFold

  2021cited by this paper
• Efficient Exploration of Chemical Space with Docking and Deep Learning.

  2021cited by this paper
• AiZynthFinder: a fast, robust and flexible open-source software for retrosynthetic planning

  2020cited by this paper
• Harnessing endophenotypes and network medicine for Alzheimer's drug repurposing

  2020cited by this paper
• KLIFS: an overhaul after the first 5 years of supporting kinase research

  2020cited by this paper
• Cognitive impairment with diabetes mellitus and metabolic disease: innovative insights with the mechanistic target of rapamycin and circadian clock gene pathways

  2019cited by this paper
• THE BLOOD-BRAIN BARRIER (BBB) SCORE.

  2019cited by this paper
• Deep learning enables rapid identification of potent DDR1 kinase inhibitors

  2019cited by this paper
• Rethinking drug design in the artificial intelligence era

  2019cited by this paper
• Circadian Rhythm and Alzheimer’s Disease

  2018cited by this paper
• Salt-inducible kinase 3 regulates the mammalian circadian clock by destabilizing PER2 protein

  2017cited by this paper
• PubChem Substance and Compound databases

  2015cited by this paper
• ff14SB: Improving the Accuracy of Protein Side Chain and Backbone Parameters from ff99SB.

  2015cited by this paper
• 'The clocks that time us'—circadian rhythms in neurodegenerative disorders

  2014cited by this paper
• Jaguar: A high-performance quantum chemistry software program with strengths in life and materials sciences

  2013cited by this paper
• Stochastic voyages into uncharted chemical space produce a representative library of all possible drug-like compounds.

  2013cited by this paper
• ChEMBL: a large-scale bioactivity database for drug discovery

  2011cited by this paper
• LEEP AND CIRCADIAN ABNORMALITIES IN A TRANSGENIC MOUSE ODEL OF ALZHEIMER ’ S DISEASE : A ROLE FOR CHOLINERGIC

  2005cited by this paper
• Glide: a new approach for rapid, accurate docking and scoring. 2. Enrichment factors in database screening.

  2004cited by this paper
• Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy.

  2004cited by this paper
• Development and testing of a general amber force field

  2004cited by this paper
• A fast SHAKE algorithm to solve distance constraint equations for small molecules in molecular dynamics simulations

  2001cited by this paper

• No citing papers are available for this paper.