Biology subjects

Woods, V.

Publications and source records attributed to Woods, V..

6 recordsLinked to original sources

bioRxiv2025

AmesNet: A Deep Learning Model Enhancing Generalization in Ames Mutagenicity Prediction

Regulatory agencies require comprehensive genotoxicity assessments for all novel small-molecule therapeutics prior to human trials. Developers often delay these studies until a candidate is nearing regulatory submission because they are expensive and secondary to bioactivity. This timing creates a bottleneck where late-stage failures can jeopardize >$10 million in capital and multiple years of developmental progress per candidate. The Ames assay is used to detect a molecules mutagenic potential. Regulators now explicitly support the use of in silico Ames mutagenicity models through enabling legislation, dedicated FDA AI toxicology programs, internationally harmonized guidelines, and benchmark challenges. However, current Ames models suffer from a dramatic sensitivity drop-off when they evaluate molecules outside their training domain. Sensitivity is the most important metric in Ames prediction because false negatives allow mutagenic compounds to advance undetected and trigger the most costly late-stage failures. Attempts to fix this sensitivity drop-off often reduce overall model performance, which can be represented by balanced accuracy. For example, DeepAmes reports high levels of sensitivity only by sacrificing its balanced accuracy. We introduce AmesNet, a novel Task-Conditioned modeling paradigm that achieves both class-leading sensitivity and balanced accuracy in novel chemical spaces. AmesNet utilizes a dual branch architecture containing a molecular encoder and a dedicated channel to condition Ames assay context such as metabolic activation and bacterial strain type. In comparative benchmarks, AmesNet reached a sensitivity of 0.73 (95% confidence interval: 0.68-0.77) and a simultaneous balanced accuracy of 0.81 (95% confidence interval: 0.79-0.83) on the out-of-domain test data. This represents an improvement in sensitivity of up to 46% over existing approaches without a trade-off in balanced accuracy. Structural analysis demonstrates that AmesNet recovers difficult-to-detect mutagenic compounds missed by existing models. This framework provides a high-confidence filtering mechanism that enables drug developers to turn a costly late-stage safety bottleneck into a proactive decision-making edge.

pharmacology and toxicology↗

bioRxiv2025

GALILEO Generatively Expands Chemical Space and Achieves One-Shot Identification of a Library of Novel, Specific, Next Generation Broad-Spectrum Antiviral Compounds at High Hit Rates

The COVID-19 pandemic (2019-2023) demonstrated the need for safe and effective, stockpiled broad-spectrum antiviral drugs to suppress unexpected viral outbreaks. The inability of the pharmaceutical industry to create such a therapeutic in the 6 years since the onset of COVID-19 demonstrates antiviral drug development must undergo a paradigm shift for this to occur. AI-based target and medicinal chemistry discovery platforms such as GALILEO and its geometric graph convolutional network tool ChemPrint, which we recently published, hold promise in accelerating and reimagining the drug development process. GALILEO identified the Thumb-1 site (an allosteric subdomain of the viral RNA polymerase) to be structurally conserved across numerous viral species and MDL-001 (an orally available therapeutic with a favorable pharmacokinetics and safety profile in humans) to be a potent inhibitor thereof. Published preclinical proof-of-concept studies demonstrated MDL-001 as a first-in-class broad-spectrum antiviral drug. This study leverages GALILEOs generative and multimodal discovery tools to create trillions of new chemical entities (NCEs) from MDL-001s pharmacophoric scaffold and select a library of highly specific and optimized compounds for next-generation broad-spectrum antiviral development. Specifically, ChemPrints one-shot predictions identified 12 NCEs with predicted affinity to Thumb-1 and significantly reduced, or no, affinity to MDL-001s original target. In vitro bioassays demonstrated a 100% hit rate, with all 12 NCEs having antiviral activity against Hepatitis C Virus (HCV) and/or human Coronavirus 229E. In vitro studies also demonstrated reduced activity of 800-fold to greater than 15,000-fold relative to MDL-001s originally designed mechanism of action (MoA). In Tanimoto similarity plots, the 12 NCEs lacked chemical relatedness to known antiviral drugs, including MDL-001 (average Tanimoto coefficient: 0.38) and Beclabuvir, (average Tanimoto coefficient: 0.13) a HCV Thumb-1 ligand that lacks broad-spectrum activity. This study showcases GALILEOs ability to generate vast NCE libraries and ChemPrints extrapolative capabilities to discover large, potent NCE libraries of compounds, specific to a complex target that are novel to known chemistry at high hit-rates.

bioinformatics↗

bioRxiv2025

MDL-001: An Oral, Safe, and Well-Tolerated Broad-Spectrum Inhibitor of Viral Polymerases

Endemic viral respiratory illnesses caused by influenza, RSV, and coronaviruses impose a global disease burden of hundreds of millions of infections and hundreds of thousands of deaths annually. Chronic hepatitis B and C infections persist in 254 million and 58 million people worldwide, respectively. No approved therapy addresses these viruses simultaneously. MDL-001 is an oral, direct-acting, broad-spectrum antiviral targeting the Thumb-1 allosteric site of viral polymerases. Here, we report MDL-001 inhibits influenza A/B, RSV, SARS-CoV-2, endemic human coronaviruses, and hepatitis B/C/D viruses with nanomolar EC90 potency in vitro. MDL-001 demonstrated equivalent in vivo efficacy to oseltamivir against mouse-lethal influenza A virus infection, reducing lung viral load by 2.6 log10 and preventing weight loss and mortality. MDL-001 demonstrated equivalent symptom reduction to subcutaneous remdesivir after SARS-CoV-2 infection. MDL-001 treatment reduced SARS-CoV-2 lung viral titers by 2.9 log10, superior to literature reported 1.4 log10 and 1.0 log10 reductions for nirmatrelvir and molnupiravir, respectively. MDL-001 reduced HCV viremia by 3.3 log10, equivalent to sofosbuvir. Oral MDL-001 reduced plasma HBV surface antigen by 2.5 log10, exceeding tenofovir alafenamide from Day 8 through Day 21. MDL-001 demonstrated an in vitro EC50 of 79.4 nM against infection with RSV strain A2, a 416-fold improvement over literature-reported ribavirin. Oral pharmacokinetics studies demonstrate MDL-001 is rapidly absorbed and partitioned into Lung (Kp=39-52) and Liver (Kp=71-104) tissues. MDL-001 produced no treatment-related adverse events across 376 animals and showed no hERG, Ames genotoxicity, or micronucleus safety issues. These findings support MDL-001 as a broad-spectrum, direct-acting non-nucleoside antiviral clinical candidate.

pharmacology and toxicology↗

bioRxiv2024

Longitudinal evidence for the emergence of multiple intelligences in assistance dog puppies

Cognitive test batteries suggest that adult dogs have different types of cognitive abilities that vary independently. In the current study, we tested puppies repeatedly over a crucial period of development to explore the timing and rate at which these different cognitive skills develop. Service dog puppies (n = 113), raised using two different socialization strategies, were either tested longitudinally (n =91) or at a single time point (n = 22). Subjects tested longitudinally participated in the battery every two weeks during and just beyond their final period of rapid brain growth (from approximately 8-20 weeks of age). Control puppies only participated in the test battery once, which allowed us to evaluate the impact of repeated testing. In support of the multiple intelligences hypothesis (MIH), cognitive skills emerged at different points across the testing period, not simultaneously. Maturational patterns also varied between cognitive skills, with puppies showing adult-like performance on some tasks only weeks after a skill emerged, while never achieving adult performance in others. Differences in rearing strategy did not lead to differences in developmental patterns while, in some cases, repeated testing did. Overall, our findings provide strong support for the MIH by demonstrating differentiated development across the cognitive abilities tested.

animal behavior and cognition↗

bioRxiv2024

Discovery of RdRp Thumb-1 as a novel broad-spectrum antiviral family of targets and MDL-001 as a potent broad-spectrum inhibitor thereof - Part I: A Bioinformatics and Deep Learning Approach

RNA viruses cause human diseases ranging from mild colds to deadly pandemics. Direct-acting, broad-spectrum, non-nucleoside antivirals have been characterized as impossible to develop because allosteric binding sites are poorly conserved. The HCV NS5B RNA-dependent RNA polymerase (RdRp) Thumb-1 allosteric site and its interaction with the HCV NS5B {Lambda}1-loop governs an essential conformational change required for polymerase initiation. The only approved NS5B Thumb-1 inhibitor, beclabuvir, has been shown to be inactive against a broad panel of non-HCV viruses, including poliovirus, rhinovirus, coronavirus, coxsackievirus, influenzavirus, and HIV. A conserved, homologous allosteric site on RdRp that spans multiple viral families has not been reported. Here, we report GALILEO's discovery that the Thumb-1 pocket, its associated {Lambda}1-loop and their interaction are conserved across RNA viral families for the first time. The discovery is validated through comparative structural analysis of Protein Data Bank (PDB) deposited viral polymerases utilizing a method that allows independent, public validation by any researcher. We further demonstrate that beclabuvir's dependence on its indole C6 carbonyl to interact with the HCV-specific residue R503 restricts its activity to HCV. We validate the target discovery with MDL-001, which does not contain a C6 carbonyl substituent. MDL-001 directly blocks viral RNA synthesis in isolated replication complexes and selects for the canonical Thumb-1 resistance mutation P495S in HCV NS5B. MDL-001 demonstrates broad-spectrum in vitro inhibition of both HCV and SARS-CoV-2. Preclinical proof of concept and development of MDL-001 across HCV, HBV, HDV, influenza, SARS-CoV-2, and RSV have been previously reported. These findings establish RdRp Thumb-1 as a conserved allosteric pocket and a druggable target for broad-spectrum direct-acting antiviral development.

bioinformatics↗

bioRxiv2024

ChemPrint: An AI-Driven Framework for Enhanced Drug Discovery

Traditional High-Throughput Screening (HTS) drug discovery is inefficient. Hit rates for compounds with clinical therapeutic potential are typically 0.5% and only up to 2% maximally. Deep learning models have enriched screening rates to 28%; however, these results include hits with non-therapeutic relevant concentrations, insufficient novelty to their training set, and traverse limited chemical space. This study introduces a novel artificial intelligence (AI)-driven platform, GALILEO, and the Molecular-Geometric Deep Learning (Mol-GDL) model, ChemPrint. This model deploys both t-distributed Stochastic Neighbor Embedding (t-SNE) data splitting to maximize chemical dissimilarity during training and adaptive molecular embeddings to enhance predictive capabilities and navigate uncharted molecular territories. When tested retrospectively, ChemPrint outperformed a panel of five models for the difficult-to-drug oncology targets, AXL and BRD4, achieving an average AUROC score of 0.897 for AXL and 0.876 for BRD4 using the t-SNE split, compared to benchmark model scores ranging from 0.826 to 0.885 for AXL and 0.801 to 0.852 for BRD4. In a zero-shot prospective study, in vitro testing demonstrated that 19 of 41 compounds nominated by ChemPrint against AXL and BRD4 demonstrated inhibitory activity at concentrations [≤] 20 {micro}M, a 46% hit rate. The 19 hits reported an average-maximum Tanimoto similarity score of 0.36 relative to their training set and scores of 0.13 (AXL) and 0.10 (BRD4) relative to clinical stage compounds for these targets. Our findings demonstrate that increasing test set difficulty through training and testing ChemPrint on datasets with maximal dissimilarity enhances the predictive capabilities of the model. This results in the discovery of compound libraries at high hit rates with low therapeutic concentrations and high chemical novelty. Taken together, the proposed platform sets a new performance standard.

bioinformatics↗