Biology subjects

Pun, F. W.

Publications and source records attributed to Pun, F. W..

4 recordsLinked to original sources

bioRxiv2026

Integrating AI and causal genetics to prioritize therapeutic targets for aging and age-related diseases

Aging is increasingly viewed as a pathologic process and a principal driver of diverse age-related diseases (ARDs). Framing aging as a disease offers an opportunity to identify therapeutic targets capable of modifying multiple chronic disorders simultaneously. Here, we developed an AI-driven target discovery framework that integrates large-scale multi-omic datasets to prioritise therapeutic targets shared between aging and 12 ARDs across four major disease areas: neurological, inflammatory, metabolic, and fibrotic disorders. We identified 29 high-confidence and 16 previously unrecognized aging-associated targets implicated across selected disease areas, together with convergent pathway perturbations characterized by robust upregulation of interferon and inflammatory signaling, alongside coordinated downregulation of MYC-driven proliferative programs, consistent with heightened inflammatory activation and reduced anabolic activity during aging. Hallmarks of aging assessment revealed chronic inflammation as the most enriched hallmark across aging and ARDs. Mendelian randomization provided genetic causal support for IL6, IL6R, NLRP3, NOS2, TLR4, and GLP1R in aging-related traits and multiple ARDs, highlighting potential opportunities for drug repurposing. Co-localization analysis further demonstrated a shared genetic signal at the IL6R locus between gene expression levels and parental survival. Together, our findings outline a scalable AI-guided multi-omic framework for identifying causal and repurposable therapeutic targets for aging and ARDs. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=70 SRC="FIGDIR/small/705676v1_ufig1.gif" ALT="Figure 1"> View larger version (32K): org.highwire.dtl.DTLVardef@b94af7org.highwire.dtl.DTLVardef@e89e80org.highwire.dtl.DTLVardef@1fde804org.highwire.dtl.DTLVardef@8b817b_HPS_FORMAT_FIGEXP M_FIG C_FIG

systems biology↗

bioRxiv2025

Advancing Target Discovery Through Disease-Specific Integration of Multi-Modal Target Identification Models and Comprehensive Target Benchmarking System

Target identification is crucial for drug development. AI-driven approaches leveraging multi-omics and computational modeling can accelerate this process. However, integrating multi-modal data for disease-specific target identification and predicting translational potential remains challenging. Moreover, the absence of a systematic evaluation framework for model performance limits confidence in target reliability. We present a unified platform combining machine learning-based target identification with comprehensive benchmarking. As a testbed, we developed Target Identification Pro (TargetPro), a disease-specific model spanning 38 diseases across oncology, metabolic, immune, fibrotic, and neurological categories. TargetPro shows strong predictive performance for clinical-stage targets and reveals disease-specific patterns, underscoring the need for tailored target detection models. We next created Target Identification Benchmark (TargetBench 1.0) to rigorously assess target identification systems, including large language models, based on their ability to recover established targets and find high-quality novel candidates. This integrated approach offers a streamlined strategy to evaluate target discovery models, ultimately improving drug development efficiency.

bioinformatics↗

bioRxiv2025

DORA AI Scientist: Multi-agent Virtual Research Team for Scientific Exploration Discovery and Automated Report Generation

Modern goal-oriented scientific research process involves hierarchical teams of researchers of diverse backgrounds performing generalist and domain-specific tasks. Many of these tasks include hypothesis generation, literature review, data collection, cleanup, processing and analysis, experimental design, virtual and physical experiments, research report and academic paper writing, reference management, bibliography and quality control. Most of these tasks can be performed automatically or in a co-pilot mode by the generative reinforcement learning systems. In this paper, we introduce a versatile multi-agent scientific exploration and draft outline research assistant (DORA), which provides multiple templates and workflows for automated or semi-automated research studies and report generation. Under user guidance, it employs hierarchical teams of AI agents based on the plug-and-play generalist and domain-specific large language models (LLMs) exploiting a variety of specialized research tools and open data repositories and generates high-quality research outputs publication drafts with maximally-accurate references. DORA is designed to minimize the time and effort required for manuscript preparation, thereby enabling researchers to devote more attention to high-value discovery tasks. The system is constantly evolving with user feedback with regular feature and resource updates. The platform is available at https://dora.insilico.com.

bioinformatics↗

bioRxiv2023

Applying Artificial Intelligence to Identify Common Targets for Treatment of Asthma, Eczema, and Food Allergy

Allergic disorders are common diseases marked by the abnormal immune response towards foreign antigens that are not pathogens. Often patients with food allergy also suffer from asthma and eczema. Given the similarities of these diseases and a shortage of effective treatments, developing novel therapeutics against common targets of multiple allergies would offer an efficient and cost-effective treatment for patients. Herein, we employed the artificial intelligence-driven target discovery platform, PandaOmics, to identify common targets for treating asthma, eczema, and food allergy. Thirty-two case-control comparisons were generated from 15, 11, and 6 transcriptomics datasets related to asthma (558 cases, 315 controls), eczema (441 cases, 371 controls), and food allergy (208 cases, 106 controls) respectively, and allocated into three meta-analyses for target identification. Top-100 high-confidence targets and Top-100 novel targets were prioritized by PandaOmics for each allergic disease. Six common high-confidence targets (i.e., IL4R, IL5, JAK1, JAK2, JAK3, and NR3C1) across all three allergic diseases have approved drugs for treating asthma and eczema. Based on the targets dysregulated expression profiles and their mechanism of action in allergic diseases, three potential therapeutic targets were proposed. IL5 was selected as a high-confidence target due to its strong involvement in allergies. PTAFR was identified for drug repurposing, while RNF19B was selected as a novel target for therapeutic innovation. Analysis of the dysregulated pathways commonly identified across asthma, eczema, and food allergy revealed the well-characterized disease signature and novel biological processes that may underlie the pathophysiology of allergies. Altogether, our study dissects the shared pathophysiology of allergic disorders and reveals the power of artificial intelligence in the exploration of novel therapeutic targets.

immunology↗