Biology subjects

Province, M.

Publications and source records attributed to Province, M..

12 recordsLinked to original sources

bioRxiv2026

Graph in Graph (GiG): A novel graph AI framework for integrating and interpreting whole-person healthmedical and omics data

Medical records and omics data are rapidly becoming standard in healthcare settings, which characterize the whole-person from dysfunctional molecules to phenotypes, and thus offer potential for precise disease diagnosis and target discovery. Whereas, it remains an open problem to systematically integrate and interprete medical record and omics data of individual patients. In this study, for the first time, we propose a novel graph AI model framework, Graph in Graph (GiG), to integrate and interpret the whole-person medical and omic datasets. Specifically, the medical record data is modeled using a person-phenotype graph, followed by omics signaling graphs of invidival patients, which enables the integration of information learned from omic-signaling graph and medical phenotype features to characterize individual patients and to prioritize important omic biomarkers and phenotypes. As an exploratory study, we applied and evaluated the GiG model to study the type 2 diabetes (T2D) and pre-T2D vs healthy using the Long Life Family Study (LLFS) cohort, which enrolls families with exceptional longevity to uncover biological mechanisms of healthy aging with medical and omics data. The evaluation results showed that GiG not only achieve a high prediction but also can interpret the prediction by ranking the essential clinical and omic biomarkers. The GiG framework can be applied to other studies by effectively integrating and interpreting medical and omic datasets for disease diagnosis and pathogenesis discovery.

bioinformatics↗

bioRxiv2026

Interpreting Omics Data Analysis with Large Language Models for Disease Target and Drug Discovery

In biomedical scientific discovery, synthesizing prior knowledge from the literature is an essential component of interpreting numerical omics data analyses for disease target identification and drug discovery. Large language models (LLMs) alone can rapidly retrieve disease mechanisms from biomedical text, but text-only outputs are general and unreliable for target and drug prioritization without cohort-specific quantitative evidence. Herein, we propose a provenance-aware Text-to-Target framework that couples schema-constrained multi-model LLM retrieval with numeric omics data analysis. The key design is a modality-aware fusion step: candidates are partitioned into overlap-supported anchors, retrieval-only hidden hubs, and network-emergent novelty nodes, then propagated into staged hypothesis and strategy generation under topology constraints. We evaluate the model in Alzheimers disease (AD) and pancreatic ductal adenocarcinoma (PDAC). In PDAC, the workflow produced a balanced 75-gene candidate universe and a 23-strategy portfolio, with significant DepMap support at both target level and strategy level. In AD, stricter candidate controls yielded a compact 34-gene universe and 14 strategies; under an expanded CRISPRbrain registry, both target-level axes were significant, with strong strategy-level enrichment. Across both diseases, final strategies preserved full provenance closure to the candidate pool, enabling end-to-end auditability from retrieval artifacts to validation outputs. These results support a transferable discovery architecture in which omics evidence constrains biological activity, LLM retrieval expands mechanistic search space, and network-aware fusion preserves interpretability. The framework provides a reproducible basis for dual-disease target prioritization and motivates continuous literature-mechanism concordance with agentic evidence-refresh loops.

bioinformatics↗

bioRxiv2025

OmniCellAgent: Towards AI Co-Scientists for Scientific Discovery in Precision Medicine

Real-world biomedical scientific discovery operates as an iterative lifecycle integrating four core pillars: the targeted identification and analysis of question-specific omics datasets; the context-aware interpretation of molecular data using rich biomedical prior knowledge database; comprehensive literature reviews; and the subjective creativity and expert intuition of human scientists. Together, these pillars drive robust evidence synthesis and novel hypothesis generation. The recently reported AI agents support automated omics analysis and literature review, which typically require users to predefine and curate disease-specific datasets, which is a process that remains challenging and time-consuming. In this study, we present OmniCellAgent, a novel multi-agent AI framework built on large-scale single-cell RNA sequencing (scRNA-seq) datasets, and can autonomously retrieve and analyze disease and control-related scRNAseq datasets of diverse cell types across tissues and conditions. Moreover, it incorporates a biomedical prior knowledge and literature review agents, and disease domain-specific expert agents to systematically annotate omic data-derived targets. By aggregating evidence across agents, the framework generates structured analytical reports and potential scientific hypotheses. We evaluated OmniCellAgent across multiple disease settings, demonstrating its ability to identify relevant datasets, generate omic data analysis results, and produce structured reports and scientific hypotheses. Our findings demonstrate that multi-agent AI systems lower the technical barriers to omics-driven research, thereby accelerating scientific discovery and hypothesis generation in biomedical research and precision medicine. The source code is publicly available at: https://github.com/FuhaiLiAiLab/OmniCellAgent.

bioinformatics↗

bioRxiv2025

Identification of novel protective loci for executive function using the trail making test part B in the Long Life Family Study

The Trail Making Test (TMT) Part B (TMT-B), a well-established assessment of cognitive function, is a frequent component of diagnostic assessments for Mild Cognitive Impairment and dementia in older adults. Identifying the genetic variants associated with the TMT-B will not only gain insights of genetic determinants of cognitive function, but also the molecular mechanisms for dementia. Published GWAS to date for TMT-B suffer from relatively low power due to the use of population level data and imputation methods. To address these deficits, we used a family-based study design to identify the genetic variants associated with the TMT-B incorporating both genome-wide linkage analysis (GWLS) and whole genome sequencing (WGS). As such, we examined the sequenced genetic determinants of TMT-B using GWLS in over 2000 participants from Long Life Family Study (LLFS). In GWLS, the estimated heritability of TMT-B was 0.29. We detected one significant linkage peak at 15q25 (LOD>3.0). Statistical fine-mapping nominated five variants including three SNPs (NTRK3-rs74031103, protective CEMIP-rs2271159, and protective AGBL1-rs4134376) and two INDELs (protective KLHL25-15:85882445:IND, and protective CEMIP-15:80893381:IND) contributing to the linkage peak. Four out of these five variants are protective for TMT-B. The rs2271159 SNP influences CEMIP expression in cerebellum and hippocampus, while the 15:80893381:IND modulates CEMIP expression in blood. Additionally, the variant rs4134376 is a basal ganglia-specific eQTL for AGBL1. In conclusion, we utilized GWLS, leveraged multi-omics data (whole genome sequence genomic data, transcriptomic data, and lipidomic data), and identified novel protective variants and genes for TMT-B performance.

genomics↗

bioRxiv2025

FISHNET: A Network-based Tool for Analyzing Gene-level P-values to Identify Significant Genes Missed by Standard Methods

FISHNET uses prior biological knowledge, represented as gene interaction networks and gene function annotations, to identify genes that do not meet the genome-wide significance threshold but replicate nonetheless. Its input is gene-level P-values from any source, including omicsWAS, aggregation of GWAS P-values, CRISPR screens, or differential expression analysis. It is based on the idea that genes whose P-values are low due to sampling error are distributed randomly across networks and functions, so genes with suggestive P-values that cluster in densely connected subnetworks and share common functions are less likely to reflect sampling error and more likely to replicate. FISHNET combines network and function analysis with permutation-based P-value thresholds to identify a small set of exceptional genes that we call FISHNET genes. Applied to 11 cardiovascular risk traits, FISHNET identified 19 gene-trait relationships that missed genome-wide significance thresholds but, nonetheless, replicated in an independent cohort. The replication rate of FISHNET genes matched or exceeded that of other genes with similar P-values. FISHNET identified a novel association between RUNX1 expression and HDL that is supported by experimental evidence that RUNX1 promotes white fat browning, which increases HDL cholesterol levels. FISHNET also identified an association between LTB expression and BMI that is supported by experimental evidence that higher LTB expression increases BMI via activation of the LT{beta}R pathway. Both associations failed genome-wide significance thresholds, highlighting FISHNETs ability to uncover meaningful relationships missed by traditional methods. FISHNET software is freely available at https://doi.org/10.5281/zenodo.14765850.

bioinformatics↗

bioRxiv2024

BioMedGraphica: An All-in-One Platform for Biomedical Prior Knowledge and Omic Signaling Graph Generation

Multi-omic data analysis is essential for scientific discovery in precision medicine. However, translating statistical results of omic data analysis into novel scientific hypothesis remains a significant challenge. Human experts must manually review analysis results and generate new hypothesis based on extensive and inter-connected biomedical prior knowledge, which is subjective and not scalable. While large language models (LLMs) can accelerate the discovery, their reasoning improves when grounded in structured, auditable and comprehensive biomedical prior knowledge. Biomedical knowledge, however, is scattered across heterogeneous databases that use diverse and inconsistent nomenclature systems, making it difficult to integrate resources into a unified format for scalable analysis. This fragmentation limits the ability of AI systems to fully leverage biomedical data for scientific discovery. To address these challenges, we developed BioMedGraphica, an all-in-one platform that harmonizes fragmented biomedical resources by integrating 11 entity types and 30 relation types from 43 databases into a unified knowledge graph containing 2,306,921 entities and 27,232,091 relations. In addition, to the best of our knowledge, this is the first work to propose a novel Textual-Numeric Graph (TNG) data-structure for multi-omics data analysis. In TNG, textual information captures prior biological knowledge (e.g., transcription start sites, functions, mechanisms), while numeric values represent quantitative biomedical features, and the integrated relations can help uncover mechanisms. By bridging prior knowledge with user-specific data, TNG is a novel and ideal data-structure for the development of graph foundation models, with the potential to improve prediction performance and interpretability, while also augmenting LLMs by supplying graph-structured mechanistic context to strengthen reasoning. The details for BioMedGraphica code can be accessed by github link: https://github.com/FuhaiLiAiLab/BioMedGraphica and BioMedGraphica knowledge graph data can be downloaded from huggingface dataset: https://huggingface.co/datasets/FuhaiLiAiLab/BioMedGraphica

bioinformatics↗

bioRxiv2024

mosGraphFlow: a novel integrative graph AI model mining disease targets from multi-omic data

Multi-omic data can better characterize complex cellular signaling pathways from multiple views compared to individual omic data. However, integrative multi-omic data analysis to rank key disease biomarkers and infer core signaling pathways remains an open problem. In this study, our novel contributions are that we developed a novel graph AI model, mosGraphFlow, for analyzing multi-omic signaling graphs (mosGraphs), 2) analyzed multi-omic mosGraph datasets of AD, and 3) identified, visualized and evaluated a set of AD associated signaling biomarkers and network. The comparison results show that the proposed model not only achieves the best classification accuracy but also identifies important AD disease biomarkers and signaling interactions. Moreover, the signaling sources are highlighted at specific omic levels to facilitate the understanding of the pathogenesis of AD. The proposed model can also be applied and expanded for other studies using multi-omic data. Model code is accessible via GitHub: https://github.com/FuhaiLiAiLab/mosGraphFlow

bioinformatics↗

bioRxiv2024

mosGraphGPT: a foundation model for multi-omic signaling graphs using generative AI

Generative pretrained models represent a significant advancement in natural language processing and computer vision, which can generate coherent and contextually relevant content based on the pre-training on large general datasets and fine-tune for specific tasks. Building foundation models using large scale omic data is promising to decode and understand the complex signaling language patterns within cells. Different from existing foundation models of omic data, we build a foundation model, mosGraphGPT, for multi-omic signaling (mos) graphs, in which the multi-omic data was integrated and interpreted using a multi-level signaling graph. The model was pretrained using multi-omic data of cancers in The Cancer Genome Atlas (TCGA), and fine-turned for multi-omic data of Alzheimers Disease (AD). The experimental evaluation results showed that the model can not only improve the disease classification accuracy, but also is interpretable by uncovering disease targets and signaling interactions. And the model code are uploaded via GitHub with link: https://github.com/mosGraph/mosGraphGPT

bioinformatics↗

bioRxiv2024

mosGraphGen: a novel tool to generate multi-omic signaling graphs to facilitate integrative and interpretable graph AI model development

Multi-omics data, i.e., genomics, epigenomics, transcriptomics, proteomics, characterize cellular complex signaling systems from multi-level and multi-view and provide a holistic view of complex cellular signaling pathways. However, it remains challenging to integrate and interpret multi-omics data for mining key disease targets and signaling pathways. Graph AI models have been widely used to analyze graph-structure datasets, and are ideal for integrative multi-omics data analysis because they can naturally integrate and represent multi-omics data as a biologically meaningful multi-level signaling graph and interpret multi-omics data via graph node and edge ranking analysis. However, it is non-trivial for graph-AI model developers to pre-analyze multi-omics data and convert the data into biologically meaningful graphs, which can be directly fed into graph-AI models. To resolve this challenge, we developed mosGraphGen (multi-omics signaling graph generator), generating Multi-omics Signaling graphs (mos-graph) of individual samples by mapping multi-omics data onto a biologically meaningful multi-level background signaling network with data normalization by aggregating measurements and aligning to the reference genome. With mosGraphGen, AI model developers can directly apply and evaluate their models using these mos-graphs. In the results, mosGraphGen was used and illustrated using two widely used multi-omics datasets of TCGA and Alzheimers disease (AD) samples. The code of mosGraphGen is open-source and publicly available via GitHub: https://github.com/FuhaiLiAiLab/mosGraphGen

systems biology↗

bioRxiv2024

PathFinder: a novel graph transformer model to infer multi-cell intra- and inter-cellular signaling pathways and communications

Recently, large-scale scRNA-seq datasets have been generated to understand the complex and poorly understood signaling mechanisms within microenvironment of Alzheimers Disease (AD), which are critical for identifying novel therapeutic targets and precision medicine. Though a set of targets have been identified, however, it remains a challenging to infer the core intra- and inter-multi-cell signaling communication networks using the scRNA-seq data, considering the complex and highly interactive background signaling network. Herein, we introduced a novel graph transformer model, PathFinder, to infer multi-cell intra- and inter-cellular signaling pathways and signaling communications among multi-cell types. Compared with existing models, the novel and unique design of PathFinder is based on the divide-and-conquer strategy, which divides the complex signaling networks into signaling paths, and then score and rank them using a novel graph transformer architecture to infer the intra- and inter-cell signaling communications. We evaluated PathFinder using scRNA-seq data of APOE4-genotype specific AD mice models and identified novel APOE4 altered intra- and inter-cell interaction networks among neurons, astrocytes, and microglia. PathFinder is a general signaling network inference model and can be applied to other omics data-driven signaling network inference.

bioinformatics↗

bioRxiv2019

Genome-wide Interaction Studies by the CHARGE Gene-Lifestyle Interactions Working Group: what we have learned and what is coming next.

The CHARGE Gene-Lifestyle Interactions Working Group is a unique initiative formed to improve our understanding of the role and biological significance of gene-environment interactions in human traits and diseases. The consortium published several multi-ancestry genome-wide interaction studies (GWIS) involving up to 610,475 individuals for three lipids and four blood pressure traits while accounting for interaction effects with drinking and smoking exposures. Here we used GWIS summary statistics from these studies to decipher potential differences in genetic associations and GxE interactions across phenotype-exposure-population trios, and to derive new insights on the potential mechanistic underlying GxE through in-silico functional analyses. Our comparative analysis shows first that interaction effects likely contribute to the commonly reported ancestry-specific genetic effect in complex traits, and second, that some phenotype-exposures pairs are more likely to benefit from a greater detection power when accounting for interactions. It also highlighted a negligible correlation between main and interaction effects, providing material for future methodological development and biological discussions. We also estimated contributions to phenotypic variance, including in particular the genetic heritability conditional on the exposure, and heritability partitioned across a range of functional annotations and cell-types. In these analyses, we found multiple instances of heterogeneity of functional partitions between exposed and unexposed individuals, providing new evidence for likely exposure-specific genetic pathways. Finally, along this work we identified potential biases in methods used to jointly meta-analyses genetic and interaction effects. We performed a series of simulations to characterize these limitations and to provide the community with guideline for future GxE studies.

genetics↗

bioRxiv2017

VarExp: Estimating variance explained by Genome-Wide GxE summary statistics

Many genomic analyses, such as genome-wide association studies (GWAS) or genome-wide screening for Gene-Environment (GxE) interactions have been performed to elucidate the underlying mechanisms of human traits and diseases. When the analyzed outcome is quantitative, the overall contribution of identified genetic variants to the outcome is often expressed as the percentage of phenotypic variance explained. In practice, this is commonly estimated using individual genotype data. However, using individual-level data faces practical and ethical challenges when the GWAS results are derived in large consortia through meta-analysis of results from multiple cohorts. In this work, we present a R package, \"VarExp\", that allows for the estimation of the percentage of phenotypic variance explained by variants of interest using summary statistics only. Our package allows for a range of models to be evaluated, including marginal genetic effects, GxE interaction effects, and main genetic and interaction effects jointly. Its implementation integrates all recent methodological developments on the topic and does not need external data to be uploaded by users.\n\nThe R source code, tutorial and associated example are available at https://gitlab.pasteur.fr/statistical-genetics/VarExp.git.

bioinformatics↗