SEARCH · Search PubMed

Results for “Machine learning.”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 199 records · Page 11Linked to original sources

PubMed2026

Metagenomic polymorphic toxin effector and immunity profiling predicts microbiome development and disease-related dysbiosis.

Bacteria use antagonistic interbacterial weapons, such as polymorphic toxin secretion systems (TSS), to compete for niches in the human gut microbiome. We hypothesized that TSS influence gut microbiome development and disease-related dysbiosis. We developed a bioinformatic marker gene approach (PolyProf) to quantify TSS including ~200 effector and immunity genes and applied it to ~15,000 publicly available human metagenomes. PolyProf alpha and beta diversity readily distinguished 12 different human disease states and enabled the construction of highly accurate linear regression classifier machine learning models. Elastic net machine learning models integrating bacterial taxonomy with PolyProf had strong predictive value for 12 disease states, outperforming models utilizing taxonomy alone. During microbiome development in the first year of life, PolyProf alpha diversity increases, and beta diversity becomes increasingly like the maternal microbiome, influenced by vertical transfer, delivery mode, and breastfeeding. PolyProf is related to strain sharing among adults through social interactions. In summary, TSS genes strongly correlate with microbiome development and interpersonal strain sharing, suggesting roles for interbacterial antagonism. Since PolyProf distinguishes diverse adult disease statuses, these dynamics may contribute to non-genetic inheritance.IMPORTANCEPrevious research has demonstrated that bacteria compete within the gut microbiome using toxin secretion systems (TSS). How TSS contribute to human microbiome development and the microbiome alterations observed in human diseases is not known. This study develops a new bioinformatic tool for profiling TSS-related genes in metagenomic data. Application of this approach to large-scale human fecal metagenomic data demonstrates the dynamic association of TSS during microbiome development, including the exchange of strains among social contacts. TSS gene abundance patterns are highly predictive of 12 disease states. This study advances the field by enabling TSS profiling in metagenomes and by identifying disease and microbiome development biomarkers that provide hypotheses for future mechanistic studies and may be useful for disease diagnosis.

Dysbiosis↗

PubMed2025

Machine learning-based analysis of the impact of 5' untranslated region on protein expression.

The 5' untranslated region (5'UTR) plays a crucial regulatory role in messenger RNA (mRNA), with modified 5'UTRs extensively utilized in vaccine production, gene therapy, etc. Nevertheless, manually optimizing 5'UTRs may encounter difficulties in balancing the effects of various cis-elements. Consequently, multiple 5'UTR libraries have been created, and machine learning models have been employed to analyze and predict translation efficiency (TE) and protein expression, providing insights into critical regulatory features. On the one hand, these screening libraries, based on TE and mean ribosome load, struggle to accurately quantify protein expression; on the other hand, a precise method for quantifying 5'UTRs necessitates a significantly costlier library. To resolve this dilemma, we constructed a library utilizing firefly luciferase as the reporter to measure accurate protein expression. In addition, we optimized the library construction method by clustering mRNA sequences to reduce redundant data and minimize the size of the dataset. This dual strategy by increasing accuracy and reducing dataset size was found to be effective in predicting the 5'UTRs from the PC3 cell line.

5' Untranslated Regions↗

PubMed2025

Comparing artificial and convolutional neural networks with traditional models for Genomic prediction in wheat.

With the rapid development of sequencing technology, the application of genomic prediction has become more and more common in breeding schemes of livestocks and crops. Selecting an appropriate statistical model is of central importance to achieve high prediction accuracy. Recently, machine learning models have been expected to upgrade genomic prediction into a new era. However, the perspective still suffers from lack of evidence that machine learning models can generally outperform the traditional ones on empirical data sets. In this study, we compared two machine learning models based on artificial neural network (ANN) and convolutional neural network (CNN) with four traditional models, including genomic best linear unbiased prediction (GBLUP), Bayesian ridge regression (BRR), BayesA and BayesB, using three published data sets for grain yield in wheat. For each model, we considered two variants: modeling and ignoring the genotype-by-environment ([Formula: see text]) interaction. In the comparison, we considered two strategies of cross-validation: predicting genotypes that have not been evaluated in any environment (CV1) and predicting genotypes that have been tested in other environments (CV2). Our results showed that traditional Bayesian models (BayesA, BayesB, and BRR) outperformed GBLUP, ANN and CNN when considering [Formula: see text] interaction. The accuracies of ANN and CNN were higher than traditional models only in CV1 and when [Formula: see text] interaction was ignored. It was also found that the performance of the two machine learning models was significantly affected by the interaction between the CV strategy and the way of treating the [Formula: see text] interaction, while that of the four traditional models was only influenced by whether the [Formula: see text] interaction was considered or not. Thus, machine learning models can be a powerful complementary to the traditional ones and their superiority may depend on the prediction scenario. Among the two machine learning models, we observed that the accuracy of ANN was higher than CNN in most cases, indicating that it is still challenging to adapt complex machine learning models such as CNN to genomic prediction.

ANN↗

PubMed2026

ALPAR: automated learning pipeline for antimicrobial resistance.

SUMMARY: The field of machine learning in antimicrobial resistance (AMR) research has experienced rapid growth, fueled by advancements in high-throughput genome sequencing and the growing capacity of computational resources. However, the complexity and lack of standardized data preparation and bioinformatic analyses present significant challenges, especially for newcomers to the domain. In response to these challenges, we introduce ALPAR (Automated Learning Pipeline for Antimicrobial Resistance), a comprehensive AMR data analysis tool covering the entire process from processing of raw genomic data to training machine learning models to interpretation of results. Our method relies on a reproducible pipeline that integrates widely used bioinformatics tools, presenting a simplified, automatic workflow specifically tailored for single-reference AMR analysis. Accepting genomic data in the form of FASTA files as input, ALPAR facilitates the generation of machine learning-ready data tables and both the training of machine learning and the execution of genome-wide association studies (GWAS) experiments. Additionally, our tool offers supplementary functionalities such as phylogeny-based analysis of the distribution of mutations, enhancing its utility for researchers. The tool has also proven its performance in competitive benchmarks, winning the 2024 CAMDA Anti-Microbial Resistance Prediction Challenge and placing third in the 2025 edition. AVAILABILITY AND IMPLEMENTATION: ALPAR is open-source and freely accessible via GitHub (https://github.com/kalininalab/ALPAR). The pipeline is fully reproducible and can be easily installed as a Conda package (https://anaconda.org/kalininalab/ALPAR).

Machine Learning↗

PubMed1996

Rule induction and instance-based learning applied in medical diagnosis.

Machine learning methods have been applied in a variety of medical domains in order to improve medical decision making. Improved medical diagnosis and prognosis can be achieved through automatic analysis of patient data stored in medical records, i.e., by learning from past experience. Given patient records with corresponding diagnoses, machine learning methods are able to classify new cases either through constructing explicit rules that generalize the training cases (e.g., rule induction) or by storing (some of) the training cases for reference (instance-based learning). This paper presents the methodologies of rule induction and instance-based learning and their application to medical diagnosis, in particular, the problem of early diagnosis of rheumatic diseases. It also discusses the possibility to use existing expert knowledge to support the learning process and the utility of such knowledge.

Algorithms↗

PubMed2026

Machine learning-based clinical tool for identifying factors associated with symptomatic knee osteoarthritis: the Nagahama study.

BACKGROUND: A clinical tool that evaluates factors associated with symptomatic knee osteoarthritis (OA) based on modifiable factors is lacking. This study aimed to develop a machine learning-based clinical assessment tool using modifiable factors to identify factors associated with symptomatic knee OA and to determine its accuracy. METHODS: This study included 429 participants (81.8% women; age, 69.0 ± 5.3 years) from the Nagahama Study who were ≥60 years old and had radiographically confirmed knee OA. A Knee Society Knee Scoring System 2011 symptom score of <23 points defined symptomatic knee OA. Participants were randomly assigned to training (70%) and test (30%) datasets. A machine learning model was developed using Extreme Gradient Boosting with 27 variables, and the SHapley Additive exPlanation (SHAP) values were used to assess feature importance. The top 8 features were translated into a 100-point clinical scoring tool weighted by their SHAP contributions. The cutoff value indicating symptomatic knee OA in the clinical assessment tool was determined using receiver operating characteristic analysis, and model performance was evaluated in both datasets. RESULTS: The clinical assessment tool consisted of low back pain, OA severity, depressive tendencies, knee flexion/extension range of motion, knee extension and hip abduction strength, and lower limb muscle quality. The model showed moderate discriminative performance (AUC 0.771 and 0.773 in the training and test datasets, respectively), with a cutoff point of 47. CONCLUSION: The proposed clinical assessment tool may provide a structured framework for assessing modifiable factors associated with symptomatic knee OA, reflecting their contribution to current symptom status.

Humans↗

PubMed2026

Unveiling the power of TIIC: A prognostic tool for esophageal adenocarcinoma.

BACKGROUND: Esophageal adenocarcinoma (EAC) remains a lethal malignancy with limited prognostic tools for guiding immunotherapy. Tumor-infiltrating immune cells (TIICs) play a critical role in EAC prognosis and treatment response. METHODS: We integrated single-cell RNA sequencing and bulk transcriptome data from TCGA and GEO databases. TIIC-specific RNAs were identified via tissue specificity index calculation combined with machine learning feature selection. Twenty machine learning algorithms were benchmarked to construct an optimal TIIC signature score (TIIC-Score) based on the comprehensive C-index. Immunotherapy response, genomic mutation, and copy number variation were analyzed. Summary-data-based Mendelian randomization (SMR) and two-sample Mendelian randomization (MR) were performed to explore genetic associations. Core prognostic TIIC-related genes were functionally validated in esophageal cancer cell lines through loss-of-function assays. RESULTS: The TIIC-Score demonstrated robust prognostic value for 1-, 2-, and 3-year overall survival across multiple cohorts, outperforming 22 published models. High TIIC-Score was associated with poor survival and increased chromosomal instability. Mutation profiling revealed high frequencies of TP53 (78.2%), TTN (48.7%), and SYNE1 (30.8%). MR analysis identified a significant association between gastro-oesophageal reflux and EAC risk at SNP rs8130507. Functionally, CCNI was upregulated in esophageal cancer cells, and its knockdown suppressed malignant phenotypes while promoting apoptosis, supporting its pro-tumorigenic role. CONCLUSION: The TIIC-Score provides a novel prognostic framework for EAC that effectively stratifies patient risk and may help identify individuals most likely to benefit from immunotherapy.

Esophageal adenocarcinoma↗

PubMed2025

MyESL: A Software for Evolutionary Sparse Learning in Molecular Phylogenetics and Genomics.

Evolutionary sparse learning uses supervised machine learning to build evolutionary models where genomic sites loci are parameters. It uses the Least Absolute Shrinkage and Selection Operator with bi-level sparsity to connect a specific phylogenetic hypothesis with sequence variation across genomic loci. The MyESL software addresses the need for open-source tools to perform evolutionary sparse learning analyses, offering features to preprocess input phylogenomic alignments, post-process output models to generate molecular evolutionary metrics, and make Least Absolute Shrinkage and Selection Operator regression adaptable and efficient for phylogenetic trees and alignments. The core of MyESL, which constructs models with logistic regressions using bi-level sparsity, is written in C++. Its input data preprocessing and result post-processing tools are developed in Python. Compared to other tools, MyESL is more computationally efficient and provides evolution-friendly inputs and outputs. These features have already enabled the use of MyESL in two phylogenomic applications, one to identify outlier sequences and fragile clades in inferred phylogenies and another to build genetic models of convergent traits. In addition to the use in a Python environment, MyESL is available as a standalone executable compatible across multiple platforms, which can be directly integrated into scripts and third-party software. The source code, executable, and documentation for MyESL are openly accessible at https://github.com/kumarlabgit/MyESL.

Phylogeny↗

PubMed2026

A machine learning-derived and functionally validated circadian rhythm signature predicts clinical outcomes and in silico drug sensitivity in colorectal cancer.

BACKGROUND: Colorectal cancer (CRC) displays considerable heterogeneity in clinical outcomes, highlighting the need for reliable prognostic biomarkers. While the aberrant expression of circadian rhythm-related genes has been implicated in cancer pathogenesis, its comprehensive role in CRC progression and predicted therapeutic vulnerabilities remains inadequately characterized. METHODS: Bulk and single-cell RNA-sequencing data were integrated from multiple CRC cohorts. A circadian rhythm signature (CRS) was developed through machine learning algorithms and validated for prognostic value. Comprehensive analyses of tumor microenvironment, genomic alterations, and drug sensitivity were performed. Furthermore, the biological function of the core gene, BHLHE40, was validated in CRC cell lines through CCK-8, EdU, and wound healing assays. RESULTS: Single-cell analysis demonstrated an elevated expression signature of circadian rhythm-related genes in dendritic cells. The optimized CRS, comprising 14 circadian rhythm-related genes, successfully categorized patients into high- and low-risk groups. Patients with a high CRS showed markedly poorer overall survival and computationally inferred immunosuppressive features, including reduced CD8+ T cell infiltration and increased M2 macrophage polarization. Genomic analysis revealed enhanced mutation burden in TP53 and alterations in RTK-RAS/WNT pathways. Notably, in vitro assays confirmed that BHLHE40 is significantly overexpressed in CRC cells. Knockdown of BHLHE40 markedly inhibited tumor cell proliferation and migration. Drug sensitivity profiling identified bexarotene and SMER-3 as potential therapeutic options for high-CRS patients. A nomogram integrating CRS with clinical parameters demonstrated superior predictive accuracy for 1-, 3-, and 5-year survival. CONCLUSIONS: The CRS represents a promising prognostic biomarker that reflects tumor immune status and genomic features, providing valuable insights for personalized treatment strategies in CRC.

Circadian rhythm↗

PubMed2026

Machine learning-based integration develops a novel lysosome-related prognostic signature associated with prognosis and immune infiltration landscape in acute myeloid leukemia.

BACKGROUND: Lysosomes are essential for intracellular degradation and recycling, and changes in their function significantly contribute to tumor growth. Nonetheless, the exact role of lysosome-related genes (LRGs) in the pathogenesis of acute myeloid leukemia (AML) is still inadequately comprehended. METHODS: Differentially expressed LRGs (DE-LRGs) between AML and control groups were identified using AML-related data extracted from the Gene Expression Omnibus (GEO). The LRGs-related prognostic genes were identified and the risk model was established using univariate COX regression analysis and machine learning algorithms, based on the data obtained from The Cancer Genome Atlas (TCGA). Subsequently, we performed comprehensive analyses regarding clinical features, functional pathways, immune microenvironment, and chemotherapeutic drugs sensitivity between the high- and low-risk groups. Reverse transcription Quantitative polymerase chain reaction (RT-qPCR) and western blot were adopted to validate the expression of prognostic genes in human bone marrow-derived cell line HS-27 A and human AML cell line MOLM-13. RESULTS: Through comprehensive analysis, a risk model was developed utilizing ten LRGs (ATP6V0E2, CALCRL, TMEM165, GZMB, HCK, TCIRG1, CD1D, GPRASP1, ABCA1, and NAGA), and this model was further validated using GEO datasets. Significant differences in clinical characteristics, functional pathways, immune microenvironment characteristics, and chemotherapeutic drug sensitivity were observed between the two risk groups In vitro validation experiment illustrated that the expression trends of ATP6V0E2, TMEM165, and ABCA1 were consistent with our bioinformatics analysis. CONCLUSION: Our study demonstrates that lysosome-associated signature might forecast the prognosis of AML patients and offer guidance for subsequent immunotherapy and chemotherapy strategies.

Acute myeloid leukemia↗

PubMed2025

Preliminary screening of urinary host protein biomarkers for Schistosomiasis haematobium: A proteome profiling study identifying candidate diagnostic targets in school-aged children.

Schistosomiasis is a major public health challenge and a globally neglected tropical disease. Schistosoma haematobium, the causative agent of urogenital schistosomiasis, is endemic in African countries; with school-aged children ages 7-15 years being the most vulnerable population. Current diagnostic methods rely on microscopy to identify parasite eggs in urine; which is labor-intensive, requires specialized skills, and often lacks sensitivity, especially in mild infections. To address these limitations, we explored host disease-related biomarkers as a promising avenue for advancing diagnosis and detection. We recruited 135 children ages 7-15 years from Zanzibar, a known transmission hotspot, and used data-independent acquisition (DIA) proteomics combined with machine learning to identify potential host protein biomarkers in urine samples from individuals infected with Schistosoma haematobium. Proteomic analysis identified 823 common host proteins in urine samples from the infected group. Machine learning algorithms highlighted candidate discriminative proteins; which were validated using enzyme-linked immunosorbent assays (ELISA). Machine learning emphasized SYNPO2, CD276, α2M, LCAT, and hnRNPM as the most discriminating biomarkers for Schistosoma haematobium infection. ELISA validation confirmed the differential expression trends of these proteins, while machine learning further validated LCAT and α2M, underscoring their diagnostic potential. Our study focused on host-derived proteins and identified key urinary protein biomarkers associated with Schistosoma haematobium infection, and offers new insights into host-parasite interactions and potential tools for non-invasive diagnostics. While validated in African pediatric populations from transmission hotspots, this host-protein approach inherently overcomes geographic limitations of parasite-based diagnostics; which is a critical advantage for surveillance in non-endemic regions where imported cases threaten gains toward elimination. These findings lay the groundwork for developing novel diagnostic approaches that could significantly improve the detection and surveillance of schistosomiasis, particularly in high-risk populations.

Humans↗

PubMed2026

Machine learning-guided risk stratification in elderly AML based on genomic, immunophenotypic and therapeutic profiles.

BACKGROUND: Elderly patients with acute myeloid leukemia (AML) exhibit considerable biological and clinical heterogeneity, hindering precise prognosis. Existing prognostic systems inadequately capture the complexity of elderly AML due to their reliance on data from younger cohorts and omission of key factors like immunophenotypic markers and therapeutic profiles. This study aimed to develop and internally validate a machine learning-based prognostic model specifically tailored to elderly AML patients. METHODS: A total of 156 patients were analyzed using a two-stage modeling strategy. Clinical and genomic variables were modeled first, followed by independent analysis of immunophenotypic features. Feature selection was performed using multilayer perceptron (MLP) and random forest (RF), while multivariate Cox regression was used for final model construction. Internal validation was conducted using 1000 bootstrap iterations to assess model stability and performance. RESULTS: The model demonstrated strong predictive performance, with a concordance index (C-index) of 0.702. Time-dependent area under the curve (AUC) and calibration plots confirmed accurate prediction of 1-, 3-, and 5-year overall survival. Decision curve analysis indicated favorable net benefit across a range of threshold probabilities. Key independent prognostic factors identified included TP53 mutations, high CD13 expression, and IDH2 mutations. CONCLUSION: This model provides a robust and interpretable tool for individualized risk stratification in elderly AML. By integrating genomic, immunophenotypic, and therapeutic variables, it may help optimize treatment decisions and improve outcomes for this vulnerable population. Future efforts should focus on external validation and integration of dynamic biomarkers.

Humans↗

PubMed2026

Machine learning-ready genomic biomarkers: ATF3 polymorphisms predict postoperative analgesic demand through AI-compatible phenotyping.

PURPOSE: To determine whether ATF3 polymorphisms can serve as genetic biomarkers for machine learning-based precision analgesia by establishing a genotype-phenotype association suitable for predictive modeling of postoperative opioid requirements. METHODS: In a prospective cohort of 167 adults undergoing abdominal surgery, ATF3 SNPs rs3122721 and rs3125293 were genotyped. A structured dataset architecture was developed to represent genetic profiles as input features for supervised learning models, enabling translational analysis of genotype‑dependent opioid consumption over 72 h. RESULTS: Patients with homozygous genotypes of the ATF3 SNPs had significantly higher opioid requirements than non‑carriers, despite reporting similar subjective pain scores. This consistent genotype‑dependent pattern provided a clinically relevant phenotype suitable for integration into predictive algorithms. CONCLUSION: ATF3 genotyping offers a promising biomarker for computationally informed precision analgesia. By linking genomic variability to clinically meaningful outcomes within a structured clinical and genomic framework, this approach supports the future development of risk-stratified clinical decision-support systems to optimize postoperative pain management.Trial registration ChiCTR1900021991, registered 30 April 2019. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at https://doi.org/10.1007/s13755-026-00480-9.

ATF3↗

PubMed1997

The predictive and explanatory power of inductive decision trees: a comparison with artificial neural network learning as applied to the noninvasive diagnosis of coronary artery disease.

BACKGROUND: This paper compares two machine learning systems, an inductive decision tree (IDT) and a back-propagation neural network (ANN), in the noninvasive assessment of coronary artery disease given a set of diagnostic input attributes. A collection of 490 patient cases were accumulated from the reference of diagnostic stress myocardial scintigraphy performed in a nuclear medicine department. All cases had correlating angiography, the results of which were used to derive the target diagnoses. Input attributes included 4 baseline clinical characteristics, 4 nonimaging stress components, and 3 scintigraphic findings. METHODS: We chose 4 possible angiographic criteria for coronary artery disease and assessed the ability of each learning system to develop a diagnostic model. The 2 machine learning systems were compared on the basis of predictive performance and explanatory power. RESULTS: Cross-validation experiments showed the 2 machine learning systems to have equivalent predictive power at the same level as the clinical scan reading. For the 70% stenosis criterion, the IDT had a sensitivity of 94 +/- 3% (mean +/- 95% confidence interval) and a specificity of 59 +/- 8%, and the ANN had a sensitivity of 97 +/- 2% and a specificity of 51 +/- 13%. However the IDT system exhibited excellent explanatory power; producing simple representations of the diagnostic models which agree with previous research. CONCLUSION: In comparison with the more widely used ANNs, the IDT learning system may bring advantages to certain problems in diagnostic classification.

Coronary Angiography↗

PubMed2026

The molecular landscape of chordoma: Current frontiers from multi-omics to artificial intelligence.

Chordoma is a rare and aggressive malignant bone tumor of the axial skeleton that has historically challenged clinicians due to its complex anatomical locations and a high recurrence rate of up to 85%. This review synthesizes the most recent advances in chordoma research and offers an overview of how multi-omics, advanced immunology, and artificial intelligence are reshaping the treatment paradigm. Central to its pathogenesis is the T-box transcription factor Brachyury, which this review highlights as both the pathognomonic diagnostic marker and the primary therapeutic vulnerability. Cutting-edge innovations targeting this driver include covalent small-molecule binders, targeted protein degradation, and peptide-centric CAR-T cells designed to attack the intracellular oncoprotein. The tumor immune microenvironment is functionally dynamic, and new dimensions in cellular therapy, such as dual-specific CAR constructs and NK-cell platforms, are being engineered to neutralize immunosuppressive factors. Beyond biological insights, the review emphasizes the role of computational biology, specifically how deep-learning and machine-learning models achieve expert-level precision in tumor segmentation and personalized survival forecasting. By integrating genomic, transcriptomic, epigenomic, and proteomic data, multiomics approaches can fully elucidate chordoma subtypes and underlying resistance mechanisms, ultimately paving the way for more precise and personalized therapeutic strategies.

Humans↗

PubMed2026

ProMeta: a meta-learning framework for robust disease diagnosis and prediction from plasma proteomics.

MOTIVATION: The plasma proteome offers a dynamic window of human health, capturing the real-time intersections between genetics and physiology. However, the application of deep learning to proteomics is currently hindered by a reliance on large-scale labeled datasets, rendering standard models ineffective for rare or novel diseases where patient samples are inherently scarce. RESULTS: Here, we present ProMeta, a meta-learning framework designed to enable robust disease modeling under extreme data restrictions. By integrating knowledge-guided pathway encoding with bi-level meta-optimization, ProMeta projects unstructured proteomic profiles into biologically interpretable functional tokens. This architecture allows the model to learn a global initialization containing transferable biological priors from biobank-scale data, facilitating rapid adaptation to novel tasks. Through comprehensive benchmark experiments, ProMeta consistently outperformed transfer learning and traditional machine learning baselines in both disease diagnosis and prediction tasks. In the most challenging 4-shot scenarios (utilizing only 2 cases and 2 controls), the model achieved robust generalization with an average AUROC of ∼0.69, representing a 24.6% relative improvement over the best-performing baseline methods. Mechanistic investigation revealed that ProMeta disentangles cases from controls in the latent space prior to task-specific adaptation, confirming the acquisition of universal biological rules rather than rote memorization. Furthermore, gradient-based interpretation identified disease-specific protein biomarkers and functional pathways consistent with known pathophysiology. Collectively, ProMeta overcomes the data-scarcity bottleneck in precision medicine, providing a scalable, interpretable framework for characterizing the full spectrum of human diseases, particularly for rare conditions lacking extensive clinical cohorts. AVAILABILITY AND IMPLEMENTATION: The source code of ProMeta is available at GitHub (https://github.com/lihan97/ProMeta).

Proteomics↗

PubMed2026

Inclusion of Multi-Omic Biomarkers Improves Prediction Accuracy of Response, Relapse, and Overall Survival in Acute Myeloid Leukemia Patients Receiving High-Intensity Induction Chemotherapy.

BACKGROUND: Despite advancements in genetic markers for acute myeloid leukemia (AML) risk stratification, outcome prediction remains challenging due to disease heterogeneity and dynamic genetic changes, highlighting the need for reliable biomarkers to improve AML treatment strategies and patient outcomes. To refine outcome predictions, we investigated the use of microbial-derived biomarkers to predict composite complete remission (CRc), relapse, and survival for patients on high- and low-intensity regimens, and to integrate those variables into the widely clinically utilized European Leukemia Network (ELN-2022) genetic risk classification model for high-intensity-treated patients. METHODS: We first developed machine learning models that integrate baseline fecal metabolomics, 16S rRNA-based stool microbiome features, and clinical metadata (sex, antibiotic administration, AML somatic mutations, and cytogenetics) from two cohorts of AML patients (n = 83) undergoing remission induction chemotherapy. Univariate tests and sparse canonical correlation analysis were employed for variable selection and to explore fecal metabolite-microbe relationships. A robust machine learning approach using XGBoost was employed, with 100 stratified data splits (80% training, 20% testing) and coarse-to-fine hyperparameter optimization. Variable importance was aggregated across all models to select key predictors. RESULTS: For high-intensity-treated patients, XGBoost models achieved aggregated AUROC scores of 0.719, 0.729, and 0.65 for CRc, relapse, and overall survival, respectively. For low-intensity-treated patients, these models achieved aggregate AUROC scores of 0.945, 0.724, and 0.768 for these same outcomes, respectively. Integrating the biomarkers identified in the high-intensity machine-learning models with the current ELN-2022 AML risk stratification system effectively stratified patients into risk categories, which obtained higher concordance indices and likelihood ratios, demonstrating improved prognostic accuracy for each outcome compared to ELN-2022 alone. CONCLUSIONS: The inclusion of microbial-derived biomarkers serves as a robust prognostic tool to improve outcome prediction in AML patients, highlighting the potential of its integration into AML risk assessment and paving the way for personalized treatment strategies and improved patient outcomes.

Humans↗

PubMed2026

Discovery and validation of a multi-protein panel for predicting non-fatal major adverse cardiovascular events in diabetic kidney disease.

OBJECTIVE: To identify plasma protein biomarkers associated with incident non-fatal major adverse cardiovascular events (MACE) in diabetic kidney disease (DKD) patients. RESEARCH DESIGN AND METHODS: We analyzed 317 DKD patients from the UK Biobank. Plasma proteomics and clinical data (demographics, metabolism, renal function) were integrated. In an exploratory discovery phase, three sequential Cox regression models (crude, socio-demographic-adjusted, socio-demographic-metabolic adjusted) screened non-fatal MACE-associated proteins. To prevent information leakage, the cohort was then randomly split into training (70%) and testing (30%) sets; machine-learning feature selection, hyperparameter optimization, and final model development were performed exclusively within the training set. The associated proteins were input into the four-step machine-learning pipeline (LASSO-Cox, random survival forest, Boruta, XGBoost-Cox). Predictive performance was validated using Kaplan-Meier survival analyses, longitudinal trajectory modeling, and ROC benchmarking. An interactive web application was deployed for clinical implementation. RESULTS: Of 1,463 plasma proteins, 561 were associated with non-fatal MACE across Cox models, with 14 overlapping proteins. Nine core proteins (ANG, IL1R1, CXCL14, ESAM, PTGDS, HAVCR1, FGFR2, IGSF8, CCL3) were validated: ANG showed the strongest non-fatal MACE association (HR = 3.88, 95%CI 2.33-6.48, p<0.001), and all high-expression groups had elevated non-fatal MACE risk. GO/KEGG enrichment highlighted inflammatory-immune pathways like positive regulation of MAPK cascade, Cytokine-cytokine receptor interaction and PI3K-Akt signaling pathway as key mechanisms. The model integrating proteins, demographic factors, and clinical variables achieved the highest predictive performance across non-fatal MACE (AUC = 0.768), myocardial infarction (MI) (0.808), and stroke (0.816) outcomes, with superior stability in cross-validation. CoxBoost + Elastic Net framework was selected as the optimal framework via benchmarking of 101 algorithms. The model demonstrated favorable calibration in high-risk patients and yielded positive net clinical benefit across decision thresholds of 5% to 45%. The web tool (https://jiangli2941.github.io/MACE-prediction-v2/) enables input of 28 variables, outputs non-fatal MACE risk status, risk probability, and highlights abnormal indicators. CONCLUSION: Plasma proteomics combined with machine learning identifies robust non-fatal MACE predictors in DKD.

Humans↗