SEARCH · Search PubMed
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.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
This article reviews machine learning methods for bioinformatics. It presents modelling methods, such as supervised classification, clustering and probabilistic graphical models for knowledge discovery, as well as deterministic and stochastic heuristics for optimization. Applications in genomics, proteomics, systems biology, evolution and text mining are also shown.
Proteins' flexibility is a feature in communicating changes in cell signaling instigated by binding with secondary messengers, such as calcium ions, associated with the coordination of muscle contraction, neurotransmitter release, and gene expression. When binding with the disordered parts of a protein, calcium ions must balance their charge states with the shape of calcium-binding proteins and their versatile pool of partners depending on the circumstances they transmit. Accurately determining the ionic charges of those ions is essential for understanding their role in such processes. However, it is unclear whether the limited experimental data available can be effectively used to train models to accurately predict the charges of calcium-binding protein variants. Here, we developed a chemistry-informed, machine-learning algorithm that implements a game theoretic approach to explain the output of a machine-learning model without the prerequisite of an excessively large database for high-performance prediction of atomic charges. We used the ab initio electronic structure data representing calcium ions and the structures of the disordered segments of calcium-binding peptides with surrounding water molecules to train several explainable models. Network theory was used to extract the topological features of atomic interactions in the structurally complex data dictated by the coordination chemistry of a calcium ion, a potent indicator of its charge state in protein. Our design created a computational tool of CaXML, which provided a framework of explainable machine learning model to annotate ionic charges of calcium ions in calcium-binding proteins in response to the chemical changes in an environment. Our framework will provide new insights into protein design for engineering functionality based on the limited size of scientific data in a genome space.
MOTIVATION: How can we identify causal genetic mechanisms governing bacterial traits? Initial efforts entrusting machine learning models to handle the task of predicting phenotype from genotype yield high accuracy scores. However, attempts to extract meaningful interpretations from the predictive models are found to be corrupted by falsely identified 'causal' features. Relying solely on pattern recognition and correlations is unreliable, significantly so in bacterial genomics settings where high-dimensionality and spurious associations are the norm. Though it is not yet clear whether we can overcome this hurdle, significant efforts are being made towards discovering potential high-risk bacterial genetic variants. In view of this, we set up open problems surrounding phenotype prediction from bacterial whole-genome datasets and extending those approaches to learning causal effects, and discuss challenges that impact the reliability of a machine's decision-making when faced with datasets of this nature. RESULTS: We identify major sources of non-injectivity in the formulation of the genotype-to-phenotype mapping function-linkage-disequilibrium, limited sampling, information loss in representations, unmeasured confounders and observational noise-and analyse their implications for machine learning applications. Using a collection of 4,140 Staphylococcus aureus isolates, we illustrate challenges surrounding the defined open problems. AVAILABILITY AND IMPLEMENTATION: Raw sequencing data are available from the European Nucleotide Archive (ENA) under project accessions ERP001012, PRJEB3174, PRJEB2655, PRJEB2756, and PRJEB2944. Assemblies and annotations were generated with the Sanger bacterial pipeline (https://github.com/sanger-pathogens/vr-codebase) and unitigs extracted using DBGWAS (https://gitlab.com/leoisl/dbgwas).
We address the problem of filtering medical news articles for targeted audiences. The approach is based on terms and one of the difficulties is extracting a feature set appropriate for the domain. This paper addresses the medical news-filtering problem using a machine learning approach. We describe the application of two supervised machine learning techniques, Decision Trees and Naïve Bayes, to automatically construct classifiers on the basis of a training set, in which news articles have been pre-classified by a medical expert and four other human readers. The goal is to classify the news articles into three groups: non-medical, medical intended for experts, and medical intended for other readers. While the general accuracy of the machine learning approach is around 78%, the accuracy of distinguishing non-medical articles from medical ones is shown to be 92%.
Recent advances in machine learning methods, along with successful applications across a wide variety of fields such as planetary science and bioinformatics, promise powerful new tools for practicing scientists. This viewpoint highlights some useful characteristics of modern machine learning methods and their relevance to scientific applications. We conclude with some speculations on near-term progress and promising directions.
PURPOSE: High glucose (HG), hypoxia (Hyp), and their combination are major risk factors for proliferative diabetic retinopathy (PDR). Although these conditions induce features of the Warburg-like metabolic reprogramming in human retinal endothelial cells (HRECs), it remains unclear whether they produce distinct metabolic and angiogenic subtypes. This study aimed to characterize the Warburg-like-associated metabolic heterogeneity induced by these PDR-related risk factors and evaluate the ability of supervised machine-learning models to distinguish these subtypes. METHODS: HRECs were cultured under normoglycemic, HG, Hyp (2% O2), and combined HG-Hyp conditions. Untargeted LC-MS/MS metabolomics quantified metabolites spanning carbohydrates, amino acids, nucleotides, and lipids. Principal component analysis (PCA) assessed overall metabolic variation, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis identified metabolic pathways associated with angiogenesis. In vitro angiogenesis assays measured endothelial tube formation and branching. Nine supervised classifiers (decision tree, logistic regression, naïve Bayes, random forest, K-Nearest Neighbors, neural network, gradient boosting, AdaBoost, and Support Vector Machine) were trained on the highest-ranked metabolites selected by the Information Gain Ratio feature-ranking approach. Model performance was evaluated using 10-fold cross-validation, leave-one-out cross-validation (LOOCV), permutation testing, and a classifier stability analysis under biologically meaningful distributional shift using an independent chemically induced hypoxia model (CoCl2). RESULTS: PCA revealed partial separation of metabolic profiles across conditions, indicating different Warburg-like metabolic subtypes. The combined HG-Hyp condition exhibited enhanced angiogenic potential relative to either HG or Hyp alone. KEGG pathway enrichment analysis identified fatty acid biosynthesis and elongation among the most significantly enriched pathways in HRECs under combined HG-Hyp conditions, alongside amino sugar and nucleotide sugar metabolism, glycerophospholipid metabolism, the pentose phosphate pathway, and glycolysis/gluconeogenesis. Supervised machine-learning classifiers distinguished these metabolic subtypes, with AdaBoost and gradient Boosting showing the most balanced, reproducible performance across 10-fold cross-validation, LOOCV, and permutation testing, and remaining the most reliable classifiers under domain-shift testing (area under the curve = 0.88, P = 0.0061). CONCLUSIONS: In this exploratory analysis, HG, Hyp, and their combination drive metabolically and functionally distinct subtypes of Warburg-like metabolic reprogramming in HRECs, with HG-Hyp in combination producing a highly angiogenic phenotype. Boosting-based ensemble classifiers provide a promising framework for detecting these subtypes even under domain-shift conditions, warranting validation in larger independent datasets. TRANSLATIONAL RELEVANCE: Integrating metabolomics with machine-learning classification offers a strategy to identify Warburg-like metabolic subtypes in retinal endothelial cells, providing insights into angiogenic mechanisms and guiding the development of targeted diagnostics or therapeutics for PDR.
BACKGROUND: Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder lacking objective biomarkers for early diagnosis. DNA methylation is a promising epigenetic marker, and machine learning offers a data-driven classification approach. However, few studies have examined whole-blood, genome-wide DNA methylation profiles for ASD diagnosis in school-aged children. METHODS: We analyzed genome-wide DNA methylation data from GEO dataset GSE113967, including 52 children with ASD and 48 typically developing (TD) controls. Differentially methylated positions (DMPs) were identified, and feature selection was performed using support vector machine-recursive feature elimination with cross-validation (SVM-RFECV). Classification models were developed using random forest (RF), extreme gradient boosting (XGBoost), and decision tree (DT) classifiers. A nomogram visualized feature contributions. RESULTS: A total of 138 DMPs differentiated ASD from TD children. Eleven CpG sites selected by SVM-RFECV formed the basis for model construction. RF and XGBoost achieved the highest accuracy (75%), with DT reaching 70%. Functional annotation indicated enrichment in cell adhesion and immune-related pathways. CONCLUSIONS: This exploratory study demonstrates the feasibility of integrating peripheral blood DNA methylation data with machine learning to distinguish children with ASD. While limited by sample size and moderate accuracy, this study provides methodological insights into the feasibility of integrating epigenetic and computational approaches for ASD-related biomarker exploration.
Microbiome sequencing measures relative rather than absolute abundances, providing no direct information about total microbial load. Normalization methods attempt to compensate, but rely on strong, often untestable assumptions that can bias inference. Experimental measurements of load (e.g., qPCR, flow cytometry) offer a solution, but remain costly and uncommon. A recent high-profile study proposed that machine learning could bypass this limitation by predicting microbial load from sequencing data alone. To evaluate this claim, we assembled mutt, the largest public database of paired sequencing and load measurements, spanning 35 studies and over 15,000 samples. Using mutt, we show that published machine learning models fail to generalize: on average they perform worse than a naive baseline that always predicted the training set mean. These failures stem from covariate shift-limited shared taxa between studies, differences in community composition, and differences in preprocessing pipelines-that silently derail model inputs. In contrast, Bayesian partially identified models do not attempt to impute microbial load, but instead propagate scale uncertainty through downstream analyses. Across 30 benchmark datasets, Bayesian partially identified models consistently outperformed normalization and machine learning approaches, providing a principled and reproducible foundation for microbiome inference.
MOTIVATION: Molecular scoring is a popular approach for studying pathway-level functional alterations with omics data. Using molecular scores for tasks such as single-sample molecular characterisation, phenotype prediction or disease stratification has several advantages compared to using omics data directly. Molecular scores provide biological interpretability and are more generalisable across datasets, facilitating data integration and machine learning applications. However, numerous scoring methods are available through different software packages, and currently there is a lack of tools to easily use these scores for model training and prediction. RESULTS: We developed pathMED, an R/Bioconductor package that unifies various scoring methods in a simple framework. Furthermore, pathMED also contains a machine learning module to train and test models that use the calculated molecular scores to predict clinical outcomes. We demonstrate some of its potential applications in three use cases using public omics data. We showed the generalisability of machine learning models trained on transcriptomic scores in predicting clinical outcomes when deploying on proteomic scores. We also demonstrated the application of transcriptomics scores in predicting breast cancer treatment response and identifying pathways strongly associated to tumour biology and treatment response. Finally, we demonstrated the benefit of integrating a novel gene set dissection step into the analysis pipeline to resolve disease heterogeneity at the pathway level. AVAILABILITY: PathMED is freely available in the Bioconductor repository (https://bioconductor.org/packages/release/bioc/html/pathMED.html). Code to reproduce the analyses is publicly available at https://github.com/GENyO-BioInformatics/pathMED_article.
With the completion of the Human Genome Project and the growing computational challenges presented by the large amount of genomic data available today, machine learning is becoming an integral part of biomedical research and plays a major role in the emerging fields of bioinformatics and computational biology. This situation offers unparalleled opportunities and unprecedented challenges to machine learning research in general and to Bayesian learning methods in particular. This paper outlines some of the opportunities and the challenges of this endeavor, it describes where the efforts of "cracking the code of life" can most benefit from a Bayesian approach, and it identifies some potential applications of Bayesian machine learning methods to the genomic analysis of squamous cell carcinomas of the head and neck.
Multivariate classification models play an increasingly important role in human factors research. In the past, these models have been based primarily on discriminant analysis and logistic regression. Models developed from machine learning research offer the human factors professional a viable alternative to these traditional statistical classification methods. To illustrate this point, two machine learning approaches--genetic programming and decision tree induction--were used to construct classification models designed to predict whether or not a student truck driver would pass his or her commercial driver license (CDL) examination. The models were developed and validated using the curriculum scores and CDL exam performances of 37 student truck drivers who had completed a 320-hr driver training course. Results indicated that the machine learning classification models were superior to discriminant analysis and logistic regression in terms of predictive accuracy. Actual or potential applications of this research include the creation of models that more accurately predict human performance outcomes.
Many attempts have been made to solve the problem of predicting protein secondary structure from the primary sequence but the best performance results are still disappointing. In this paper, the use of a machine learning algorithm which allows relational descriptions is shown to lead to improved performance. The Inductive Logic Programming computer program, Golem, was applied to learning secondary structure prediction rules for alpha/alpha domain type proteins. The input to the program consisted of 12 non-homologous proteins (1612 residues) of known structure, together with a background knowledge describing the chemical and physical properties of the residues. Golem learned a small set of rules that predict which residues are part of the alpha-helices--based on their positional relationships and chemical and physical properties. The rules were tested on four independent non-homologous proteins (416 residues) giving an accuracy of 81% (+/- 2%). This is an improvement, on identical data, over the previously reported result of 73% by King and Sternberg (1990, J. Mol. Biol., 216, 441-457) using the machine learning program PROMIS, and of 72% using the standard Garnier-Osguthorpe-Robson method. The best previously reported result in the literature for the alpha/alpha domain type is 76%, achieved using a neural net approach. Machine learning also has the advantage over neural network and statistical methods in producing more understandable results.
The machine learning program GOLEM was applied to discover topological rules in the packing of beta-sheets in alpha/beta-domain proteins. Rules (constraints) were determined for four features of beta-sheet packing: (i) whether a beta-strand is at an edge; (ii) whether two consecutive beta-strands pack parallel or anti-parallel; (iii) whether two beta-strands pack adjacently; and (iv) the winding direction of two consecutive beta-strands. Rules were found with high predictive accuracy and coverage. The errors were generally associated with complications in domain folds, especially in one doubly would domains. Investigation of the rules revealed interesting patterns, some of which were known previously, others that are novel. Novel features include (i) the relationship between pairs of sequential strands is in general one of decreasing size; (ii) more sequential pairs of strands wind in the direction out than in; and (iii) it takes a larger alteration in hydrophobicity to change a strand from winding in the direction out than in. These patterns in the data may be the result of folding pathways in the domains. The rules found are of predictive value and could be used in the combinatorial prediction of protein structure, or as a general test of model structures, e.g. those produced by threading. We conclude that machine learning has a useful role in the analysis of protein structures.
Sweet osmanthus (Osmanthus fragrans), a traditional ornamental plant in China, exhibits substantial variation in autumn flowering time, which significantly affects landscape application and cultivation efficiency. Here, we performed a genome-wide association study on 127 resequenced accessions classified into early, intermediate, and late flowering types, using a set of 2,325,410 single-nucleotide polymorphisms (SNPs) and 246,824 structural variants (SVs). By integrating SNP/insertion and deletion (Indel) and SV data with weighted gene co-expression network analysis, machine learning, and genomic prediction, we dissected the genetic architecture of flowering time. We identified 24 associated SNP/Indels and six SVs, mapping to 30 candidate genes, including known flowering regulators FLK, LOS1, Y14, MIF2, and GID1B. These genes showed tissue-specific expression, with some responding to low temperature. The two hub genes, GUX1 and LYG027904, were located within modules of the co-expression network associated with low-temperature treatment. Haplotype analysis revealed a specific three-SNP haplotype associated with late flowering and linked to LOS1, and epistatic interactions among combined genotypes contributed to phenotypic variation. Notably, integrating SVs with SNP/Indels improved genomic prediction accuracy; the gradient boosting decision tree model outperformed other machine learning algorithms, achieving a mean accuracy of 0.859 and an AUC > 0.8 (where AUC is area under receiver operating characteristic curve) for all flowering types. These findings provide insights into the genetic mechanisms underlying flowering time variation in O. fragrans, offer candidate genes and haplotypes for molecular breeding, and highlight the value of integrating SVs with machine learning for genomic prediction in woody ornamentals.
BACKGROUND: Myasthenia gravis (MG) is a prototypical antibody-mediated autoimmune disease with variable treatment responses with a need for biomarkers to guide therapeutic decision making. Proteomic profiling, coupled with machine learning, offers a hypothesis-free approach to identify multi-protein signatures associated with treatment response. METHODS: We analyzed sera collected at entry (baseline) from participants in a phase 3 trial randomized trial comparing thymectomy plus prednisone versus prednisone alone, along with matched controls using liquid chromatography-mass spectrometry. We derived disease-specific proteomic signatures and evaluated associations between baseline proteins and 6-month clinical outcomes using multiple machine-learning approaches with internal validation. RESULTS: Baseline serum proteomes distinguished MG from controls, with pathway enrichment implicating complement activation, immunoglobulin production, and T-cell receptor signaling. Distinct protein panels predicted 6-month clinical improvement within each treatment arm. In the thymectomy-plus-prednisone group, models captured non-linear relationships of predictive proteins in contrast with the predominant additive patterns observed in the prednisone-alone group. Predictive proteins were enriched for T-cell signaling and leukocyte trafficking functions, providing insight into treatment-specific biology. CONCLUSIONS: Baseline serum proteomics captures core disease characteristics of MG and predicts short-term clinical response in a treatment-specific manner. While our results require validation in independent cohorts, these findings could enable biomarker-guided selection of thymectomy, refine risk stratification, and furnish mechanistic readouts for future MG trials and clinical care. We aim to conduct future studies using -omic approaches to validate these baseline predictive biomarkers and pathways of treatment response in patients with MG.
The use of modern abductive machine learning techniques is described for modeling and predicting outcome parameters in terms of input parameters in medical survey data. The AIM (Abductory Induction Mechanism) abductive network machine-learning tool is used to model the educational score in a health survey of 2,720 Albanian primary school children. Data included the child's age, gender, vision, nourishment, parasite infection, family size, parents' education, and educational score. Models synthesized by training on just 100 cases predict the educational score output for the remaining 2,620 cases with 100% accuracy. Simple models represented as analytical functions highlight global relationships and trends in the survey population. Models generated are quite robust, with no change in the basic model structure for a 10-fold increase in the size of the training set. Compared to other statistical and neural network approaches, AIM provides faster and highly automated model synthesis, requiring little or no user intervention.
The paper presents machine learning (ML) models that predict sedimentation in the harbour basin of the Port of Rotterdam. The important factors affecting the sedimentation process such as waves, wind, tides, surge, river discharge, etc. are studied, the corresponding time series data is analysed, missing values are estimated and the most important variables behind the process are chosen as the inputs. Two ML methods are used: MLP ANN and M5 model tree. The latter is a collection of piece-wise linear regression models, each being an expert for a particular region of the input space. The models are trained on the data collected during 1992-1998 and tested by the data of 1999-2000. The predictive accuracy of the models is found to be adequate for the potential use in the operational decision making.
OBJECTIVES: The aim was to evaluate the potential for monotonicity constraints to bias machine learning systems to learn rules that were both accurate and meaningful. METHODS: Two data sets, taken from problems as diverse as screening for dementia and assessing the risk of mental retardation, were collected and a rule learning system, with and without monotonicity constraints, was run on each. The rules were shown to experts, who were asked how willing they would be to use such rules in practice. The accuracy of the rules was also evaluated. RESULTS: Rules learned with monotonicity constraints were at least as accurate as rules learned without such constraints. Experts were, on average, more willing to use the rules learned with the monotonicity constraints. CONCLUSIONS: The analysis of medical databases has the potential of improving patient outcomes and/or lowering the cost of health care delivery. Various techniques, from statistics, pattern recognition, machine learning, and neural networks, have been proposed to "mine" this data by uncovering patterns that may be used to guide decision making. This study suggests cognitive factors make learned models coherent and, therefore, credible to experts. One factor that influences the acceptance of learned models is consistency with existing medical knowledge.