Opal Therapeutics is expanding its women’s health discovery platform with a molecular modeling program focused on identifying and optimizing non-hormonal therapeutics. Built alongside our patient-derived uterine organoids and AI imaging tools, this in-progress effort will enable virtual screening and structure-based drug design for gynecological conditions like endometriosis and fibroids.

Endometriosis and alternative sources of non-hormonal contraception are neglected and challenging issues associated with women's health. Today, I will discuss how we leverage multiomics data and AI to identify novel targets implicated in Endometriosis and how we contribute to the challenge of designing novel non-hormonal contraceptives using AI.

PCOS impacts 1 in 10 women of reproductive age, yet has no specifically-approved therapeutics. In this heterogeneous endocrine disorder, androgen excess is linked to reproductive dysfunction and hirsutism. In our pursuit of novel therapies for PCOS, we have successfully applied our machine learning–driven medicinal chemistry platform to rapidly optimize a series of small molecules which affect the testosterone biosynthesis pathway, demonstrating in vivo reduction of testosterone with advanced leads.

We identified non-hallucinogenic 5-HT2AR agonists with antidepressant-like activity through AI-driven drug design. These molecules showed strong in vitro 5-HT2AR activation, high brain penetration in rodents, and antidepressant-like effects in behavioral and EEG tests without hallucinogenic responses.

Discovering mRNA-targeted small molecule drugs presents challenges in identifying optimal targets and developing potent, specific modulators. This presentation will explore how advanced experimental tools and computational techniques, including AI, integrated within our ibVIS platform can enhance target identification, screening, hit-to-lead, and lead optimization. New data and effective strategies will be shared to advance drug discovery programs while avoiding potential pitfalls.

Lead discovery is shifting toward hard-to-drug targets, while fast-growing chemical spaces offer new hit-finding opportunities. Yet, technologies for exploiting vast spaces to identify leads against challenging targets are yet to emerge. We present our effort on addressing these challenges by enhancing our FRASE-bot platform to include 3D pharmacophore searches on multi-billion datasets, ABFE simulations, and new strategies for extracting from phenotypic data, with a focus on lead identification and optimization.

Many companies—including Iambic—have worked on laboratory automation to generate large volumes of high-quality laboratory data. But in drug discovery it’s more important to have good solutions for many challenges than perfection for one. At Iambic we have built Enchant to address this issue head-on: Enchant is a multimodal transformer trained on dozens of modalities and drug-discovery data sources. As such it can be deployed on a huge range of highly relevant issues in drug discovery. Here we’ll discuss what it took to create Enchant, and how we leverage this and other AI technologies in our drug discovery pipeline.

Collaborations Pharmaceuticals, Inc. is a small drug discovery company that has developed its own suite of machine learning tools. These technologies have been applied to various drug discovery and toxicology targets to build and validate models prior to screening drug or compound libraries and eventual in vitro testing. Recent applications include Glycogen Synthase Kinase 3 beta (GSK-3ß or GSK-3B) as well as Chemokine receptors CCR3, CCR4, and CCR5. We will describe how these and other examples can be used to illustrate how we use in vitro and machine learning to identify new molecules for multiple diseases impacting human health.

Inhibition of MALT1 may have benefit for hematological malignancies where MALT1 is constitutively activated, such as activated B-cell (ABC)-DLBCL, as a single agent or in combination with BCR signaling pathway modulators such as BTK inhibitors. Design of novel allosteric MALT1 inhibitors using MD, ML and AI generative design approaches will be described, leading to REC-3565 development candidate, with selectivity over UGT1A1, which distinguishes it from other MALT1 inhibitors in clinical development.