Biology subjects

MA, T.

Publications and source records attributed to MA, T..

5 recordsLinked to original sources

bioRxiv2026

Trustworthy ML/AI for Aging Clocks: Preventing Systematic Prediction Bias in Biological Age Estimation

Machine learning (ML)- and artificial intelligence (AI)-based aging clocks are increasingly used to quantify physiological and molecular aging from omics and medical imaging data as distinct from chronological age. Here, we characterize a fundamental but underappreciated statistical limitation of commonly used ML/AI regression models for continuous outcomes: systematic prediction bias and its propagation to downstream association estimates. This issue becomes more challenging when the true outcome, biological age, is latent and therefore unobserved during ML/AI model training. We demonstrate that systematic prediction bias can distort and, in some cases, even reverse downstream association analyses that use aging clocks as ML/AI-predicted outcomes to assess their associations with exposures or clinical factors. For example, it can produce spurious associations suggesting that older predicted brain age is linked to better cognitive performance, or that older epigenetic age is associated with better kidney function. To address this problem, we introduce a principled and broadly applicable ML/AI regression framework based on constrained optimization, yielding better calibrated aging-clock estimates and valid downstream inference.

bioinformatics↗

bioRxiv2025

Inferring non-coding RNA regulatory network from transcriptomic data and curated databases

Non-coding RNAs (ncRNAs) have played an indispensable role in regulating gene expression and cellular processes. However, the complex regulatory circuits between ncRNAs and coding genes remain understudied. Major challenges for inferring ncRNA regulatory network (NRN) on a transcriptome-wide scale include high dimensionality of both ncRNA and coding gene expression data, their context-dependent interaction, and lack of validation. We hereby propose a comprehensive analytical framework, namely Construction and Analysis of non-coding RNA regulatory NETwork (CAR-NET), specifically designed to infer NRN from transcriptomic data and curated databases on experimentally validated ncRNA-gene interactions and disease-related ncRNAs. At its core is a novel computationally efficient Bayesian network structure learning method that leverages the semi-bipartite graph structure of NRN for dimension reduction, edge orientation and global network search. We also developed and applied algorithms to perform differential network, subnetwork and pathway analysis of the inferred network. We showed the strength of CAR-NET in simulations and three case studies, encompassing lncRNA regulation in brain development, miRNA differential regulation in renal cell carcinoma and regulation by different classes of ncRNAs in different cell lines using bulk or single cell RNA-seq data. An R-shiny app (https://github.com/kehongjie/CAR-NET) was developed to implement the method(s) with interactive GUI.

bioinformatics↗

bioRxiv2024

Orchestrating the Acquisition of Oligodendrocyte Precursor Cell versus Olfactory Bulb Interneuron Fates through Olig1/2 during Mammalian Cortical Gliogenesis and Gliomagenesis

The hijacking of developmental gliogenesis programs is a hallmark of glioblastoma (GBM), in which glial precursor cells (GPCs) typically differentiate into both neurons and glial cells. In GBM, this process is disrupted, leading to the overproduction of proliferative glial-like cells. Our study demonstrates that the knockout of Olig1/2 in both normal development and gliomagenesis causes GPCs to shift from generating highly proliferative oligodendrocyte precursor cells to producing non-proliferative olfactory bulb interneurons. Mechanistically, Olig1/2 play dual roles by orchestrating distinct transcriptional programs in GPCs, particularly inhibiting the expression of Gsx2 through direct binding to its multiple enhancers. Additionally, we provide compelling evidence that human H3.3G34R/V-mutant tumors, a subtype of high-grade gliomas, originate from dorsal cortical-derived GPCs rather than from the previously assumed progenitors in the ventral basal ganglia. Collectively, our findings reveal a previously unrecognized role of Olig1/2 in both gliogenesis and gliomagenesis, offering deeper insights into the connections between normal neural development and tumorigenesis. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=159 HEIGHT=200 SRC="FIGDIR/small/623106v1_ufig1.gif" ALT="Figure 1"> View larger version (69K): org.highwire.dtl.DTLVardef@10ea118org.highwire.dtl.DTLVardef@1725006org.highwire.dtl.DTLVardef@1e3ce25org.highwire.dtl.DTLVardef@f300ee_HPS_FORMAT_FIGEXP M_FIG C_FIG

neuroscience↗

bioRxiv2023

APOE4 poses opposite effects of plasma LDL on white matter integrity in older adults

INTRODUCTIONAPOE4 is a strong genetic risk factor of Alzheimers disease and is associated with changes in metabolism. However, the interactive relationship between APOE4 and plasma metabolites on the brain remains largely unknown. MEHODSIn the UK Biobank, we investigated the moderation effects of APOE4 on the relationship between 249 plasma metabolites derived from nuclear magnetic resonance spectroscopy on whole-brain white matter integrity, measured by fractional anisotropy using diffusion magnetic resonance imaging. RESULTSThe increase in the concentration of metabolites, mainly LDL and VLDL, is associated with a decrease in white matter integrity (b= -0.12, CI= [-0.14, -0.10]) among older APOE4 carriers, whereas an increase (b= 0.05, CI= [0.04, 0.07]) among non-carriers, implying a significant moderation effect of APOE4 (b= -0.18, CI= [-0.20,-0.15]). DISCUSSIONThe results suggest that lipid metabolism functions differently in APOE4 carriers compared to non-carriers, which may inform the development of targeted interventions for APOE4 carriers to mitigate cognitive decline.

neuroscience↗

bioRxiv2020

Genetic Fine-mapping with Dense Linkage Disequilibrium Blocks: genetics of nicotine dependence

Fine-mapping is an analytical step to perform causal prioritization of the polymorphic variants on a trait-associated genomic region observed from genome-wide association studies (GWAS). The prioritization of causal variants can be challenging due to the linkage disequilibrium (LD) patterns among hundreds to thousands of polymorphisms associated with a trait. We propose a novel{ell} 0 graph norm shrinkage algorithm to select causal variants from dense LD blocks consisting of highly correlated SNPs that may not be proximal or contiguous. We extract dense LD blocks and perform regression shrinkage to calculate a prioritization score to select a parsimonious set of causal variants. Our approach is computationally efficient and allows performing fine-mapping on thousands of polymorphisms. We demonstrate its application using a large UK Biobank (UKBB) sample related to nicotine addiction. Our results suggest that polymorphic variances in both neighboring and distant variants can be consolidated into dense blocks of highly correlated loci. Simulations were used to evaluate and compare the performance of our method and existing fine-mapping algorithms. The results demonstrated that our method outperformed comparable fine-mapping methods with increased sensitivity and reduced false-positive error rate regarding causal variant selection. The application of this method to smoking severity trait in UKBB sample replicated previously reported loci and suggested the causal prioritization of genetic effects on nicotine dependency. Author summaryDisentangling the complex linkage disequilibrium (LD) pattern and selecting the underlying causal variants have been a long-term challenge for genetic fine-mapping. We find that the LD pattern within GWAS loci is intrinsically organized in delicate graph topological structures, which can be effectively learned by our novel{ell} 0 graph norm shrinkage algorithm. The extracted LD graph structure is critical for causal variant selection. Moreover, our method is less constrained by the width of GWAS loci and thus can fine-map a massive number of correlated SNPs.

genetics↗