Machine Learning Reduced Gene/Non-Coding RNA Features That Classify Schizophrenia Patients Accurately and Highlight Insightful Gene Clusters.
Ontology highlight
ABSTRACT:
SUBMITTER: Liu Y
PROVIDER: S-EPMC8037538 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Microbial communities are important to human health. Bacterial vaginosis (BV) is a disease associated with the vagina microbiome. While the causes of BV are unknown, the microbial community in the vagina appears to play a role. We use three different machine-learning techniques to classify microbial communities into BV categories. These three techniques include genetic programming (GP), random forests (RF), and logistic regression (LR). We evaluate the classification accuracy of each of these techniques on two different datasets. We then deconstruct the classification models to identify important features of the microbial community. We found that the classification models produced by the machine learning techniques obtained accuracies above 90% for Nugent score BV and above 80% for Amsel criteria BV. While the classification models identify largely different sets of important features, the shared features often agree with past research.
| S-EPMC3912131 | biostudies-literature
Project description:Thioamide substitutions of the peptide backbone have been shown to stabilize therapeutic and imaging peptides toward proteolysis. In order to rationally design thioamide modifications, we have developed a novel Rosetta custom score function to classify thioamide positional effects on proteolysis in substrates of serine and cysteine proteases. Peptides of interest were docked into proteases using the FlexPepDock application in Rosetta. Docked complexes were modified to contain thioamides parametrized through the creation of custom atom types in Rosetta based on ab intio simulations. Thioamide complexes were simulated, and the resultant structural complexes provided features for machine learning classification as the decomposed values of the Rosetta score function. An ensemble, majority voting model was developed to be a robust predictor of previously unpublished thioamide proteolysis holdout data. Theoretical control simulations with pseudo-atoms that modulate only one physical characteristic of the thioamide show differential effects on prediction accuracy by the optimized voting classification model. These pseudo-atom model simulations, as well as statistical analyses of the full thioamide simulations, implicate steric effects on peptide binding as being primarily responsible for thioamide positional effects on proteolytic resistance.
| S-EPMC7759720 | biostudies-literature
Project description:Determining the cancer type and molecular subtype has important clinical implications. The primary site is however unknown for some malignancies discovered in the metastatic stage. Moreover liquid biopsies may be used to screen for tumoral DNA, which upon detection needs to be assigned to a site-of-origin. Classifiers based on genomic features are a promising approach to prioritize the tumor anatomical site, type and subtype. We examined the predictive ability of causal (driver) somatic mutations in this task, comparing it against global patterns of non-selected (passenger) mutations, including features based on regional mutation density (RMD). In the task of distinguishing 18 cancer types, the driver mutations-mutated oncogenes or tumor suppressors, pathways and hotspots-classified 36% of the patients to the correct cancer type. In contrast, the features based on passenger mutations did so at 92% accuracy, with similar contribution from the RMD and the trinucleotide mutation spectra. The RMD and the spectra covered distinct sets of patients with predictions. In particular, introducing the RMD features into a combined classification model increased the fraction of diagnosed patients by 50 percentage points (at 20% FDR). Furthermore, RMD was able to discriminate molecular subtypes and/or anatomical site of six major cancers. The advantage of passenger mutations was upheld under high rates of false negative mutation calls and with exome sequencing, even though overall accuracy decreased. We suggest whole genome sequencing is valuable for classifying tumors because it captures global patterns emanating from mutational processes, which are informative of the underlying tumor biology.
| S-EPMC6483366 | biostudies-literature
Project description:Gene expression is controlled by the concerted interactions between transcription factors and chromatin regulators. While recent studies have identified global chromatin state changes across cell-types, it remains unclear to what extent these changes are co-regulated during cell-differentiation. Here we present a comprehensive computational analysis by assembling a large dataset containing genome-wide occupancy information of 5 histone modifications in 27 human cell lines (including 24 normal and 3 cancer cell lines) obtained from the public domain, followed by independent analysis at three different representations. We classified the differentiation stage of a cell-type based on its genome-wide pattern of chromatin states, and found that our method was able to identify normal cell lines with nearly 100% accuracy. We then applied our model to classify the cancer cell lines and found that each can be unequivocally classified as differentiated cells. The differences can be in part explained by the differential activities of three regulatory modules associated with embryonic stem cells. We also found that the "hotspot" genes, whose chromatin states change dynamically in accordance to the differentiation stage, are not randomly distributed across the genome but tend to be embedded in multi-gene chromatin domains, and that specialized gene clusters tend to be embedded in stably occupied domains.
| S-EPMC3282719 | biostudies-literature
Project description:Investigators developed an online educational module (ESTIMATE) to teach Gastroenterology (GI) trainees how to estimate polyp size using a snare. Key components include video instruction and real-time feedback incorporated over a 40-item polyp size assessment test. Trainees from GI fellowship programs will be randomized to one of four groups: control (no video, no feedback), video-only, feedback-only, and video + feedback. Participants will classify polyps into one of three size categories:- diminutive (1-5 mm), small (6-9 mm), and large (≥10 mm). Primary outcome is accuracy of polyp size classification [diminutive (1-5 mm), small (6-9 mm), and large (≥10 mm)]. Secondary outcomes include accuracy of exact polyp size (in mm), cumulative accuracy (to plot learning curves), confidence level of polyp size classification, and directionality of inaccuracy (polyp size overestimation vs underestimation).
| 59125 | ecrin-mdr-crc
Project description:Background and purpose: The World Health Organization has recently placed new emphasis on the integration of genetic information for gliomas. While tissue sampling remains the criterion standard, noninvasive imaging techniques may provide complimentary insight into clinically relevant genetic mutations. Our aim was to train a convolutional neural network to independently predict underlying molecular genetic mutation status in gliomas with high accuracy and identify the most predictive imaging features for each mutation.Materials and methods: MR imaging data and molecular information were retrospectively obtained from The Cancer Imaging Archives for 259 patients with either low- or high-grade gliomas. A convolutional neural network was trained to classify isocitrate dehydrogenase 1 (IDH1) mutation status, 1p/19q codeletion, and O6-methylguanine-DNA methyltransferase (MGMT) promotor methylation status. Principal component analysis of the final convolutional neural network layer was used to extract the key imaging features critical for successful classification.Results: Classification had high accuracy: IDH1 mutation status, 94%; 1p/19q codeletion, 92%; and MGMT promotor methylation status, 83%. Each genetic category was also associated with distinctive imaging features such as definition of tumor margins, T1 and FLAIR suppression, extent of edema, extent of necrosis, and textural features.Conclusions: Our results indicate that for The Cancer Imaging Archives dataset, machine-learning approaches allow classification of individual genetic mutations of both low- and high-grade gliomas. We show that relevant MR imaging features acquired from an added dimensionality-reduction technique demonstrate that neural networks are capable of learning key imaging components without prior feature selection or human-directed training.
| S-EPMC6880932 | biostudies-literature
Project description:Past and recent attempts at devising objective biomarkers for traumatic brain injury (TBI) in both blood and cerebrospinal fluid have focused on abundance measures of time-dependent proteins. Similar independent determinants would be most welcome in diagnosing the most common form of TBI, mild TBI (mTBI), which remains difficult to define and confirm based solely on clinical criteria. There are currently no consensus diagnostic measures that objectively define individuals as having sustained an acute mTBI. Plasma metabolomic analyses have recently evolved to offer an alternative to proteomic analyses, offering an orthogonal diagnostic measure to what is currently available. The purpose of this study was to determine whether a developed set of metabolomic biomarkers is able to objectively classify college athletes sustaining mTBI from non-injured teammates, within 6 hours of trauma and whether such a biomarker panel could be effectively applied to an independent cohort of TBI and control subjects. A 6-metabolite panel was developed from biomarkers that had their identities confirmed using tandem mass spectrometry (MS/MS) in our Athlete cohort. These biomarkers were defined at ?6 hours following mTBI and objectively classified mTBI athletes from teammate controls, and provided similar classification of these groups at the 2, 3, and 7 days post-mTBI. The same 6-metabolite panel, when applied to a separate, independent cohort provided statistically similar results despite major differences between the two cohorts. Our confirmed plasma biomarker panel objectively classifies acute mTBI cases from controls within 6 hours of injury in our two independent cohorts. While encouraged by our initial results, we expect future studies to expand on these initial observations.
| S-EPMC5909890 | biostudies-literature
Project description:To evaluate evidence for de novo etiologies in schizophrenia, we sequenced at high coverage the exomes of families recruited from two populations with distinct demographic structures and history. We sequenced a total of 795 exomes from 231 parent-proband trios enriched for sporadic schizophrenia cases, as well as 34 unaffected trios. We observed in cases an excess of de novo nonsynonymous single-nucleotide variants as well as a higher prevalence of gene-disruptive de novo mutations relative to controls. We found four genes (LAMA2, DPYD, TRRAP and VPS39) affected by recurrent de novo events within or across the two populations, which is unlikely to have occurred by chance. We show that de novo mutations affect genes with diverse functions and developmental profiles, but we also find a substantial contribution of mutations in genes with higher expression in early fetal life. Our results help define the genomic and neural architecture of schizophrenia.
| S-EPMC3556813 | biostudies-literature
Project description:BackgroundAutomatic cell type identification is essential to alleviate a key bottleneck in scRNA-seq data analysis. While most existing classification tools show good sensitivity and specificity, they often fail to adequately not-classify cells that are missing in the used reference. Additionally, many tools do not scale to the continuously increasing size of current scRNA-seq datasets. Therefore, additional tools are needed to solve these challenges.ResultsscAnnotatR is a novel R package that provides a complete framework to classify cells in scRNA-seq datasets using pre-trained classifiers. It supports both Seurat and Bioconductor's SingleCellExperiment and is thereby compatible with the vast majority of R-based analysis workflows. scAnnotatR uses hierarchically organised SVMs to distinguish a specific cell type versus all others. It shows comparable or even superior accuracy, sensitivity and specificity compared to existing tools while being able to not-classify unknown cell types. Moreover, scAnnotatR is the only of the best performing tools able to process datasets containing more than 600,000 cells.ConclusionsscAnnotatR is freely available on GitHub ( https://github.com/grisslab/scAnnotatR ) and through Bioconductor (from version 3.14). It is consistently among the best performing tools in terms of classification accuracy while scaling to the largest datasets.
| S-EPMC8762856 | biostudies-literature
Project description:ObjectivesTo compare single parameter thresholding with multivariable probabilistic classification of ischemic stroke regions in the analysis of computed tomography perfusion (CTP) parameter maps.MethodsPatients were included from two multicenter trials and were divided into two groups based on their modified arterial occlusive lesion grade. CTP parameter maps were generated with three methods-a commercial method (ISP), block-circulant singular value decomposition (bSVD), and non-linear regression (NLR). Follow-up non-contrast CT defined the follow-up infarct region. Conventional thresholds for individual parameter maps were established with a receiver operating characteristic curve analysis. Probabilistic classification was carried out with a logistic regression model combining the available CTP parameters into a single probability.ResultsA total of 225 CTP data sets were included, divided into a group of 166 patients with successful recanalization and 59 with persistent occlusion. The precision and recall of the CTP parameters were lower individually than when combined into a probability. The median difference [interquartile range] in mL between the estimated and follow-up infarct volume was 29/23/23 [52/50/52] (ISP/bSVD/NLR) for conventional thresholding and was 4/6/11 [31/25/30] (ISP/bSVD/NLR) for the probabilistic classification.ConclusionsMultivariable probability maps outperform thresholded CTP parameter maps in estimating the infarct lesion as observed on follow-up non-contrast CT. A multivariable probabilistic approach may harmonize the classification of ischemic stroke regions.Key points• Combining CTP parameters with a logistic regression model increases the precision and recall in estimating ischemic stroke regions. • Volumes following from a probabilistic analysis predict follow-up infarct volumes better than volumes following from a threshold-based analysis. • A multivariable probabilistic approach may harmonize the classification of ischemic stroke regions.
| S-EPMC9381605 | biostudies-literature