MindReader: Unsupervised Classification of Electroencephalographic Data.

Electroencephalogram (EEG) interpretation plays a critical role in the clinical assessment of neurological conditions, most notably epilepsy. However, EEG recordings are typically analyzed manually by highly specialized and heavily trained personnel. Moreover, the low rate of capturing abnormal events during the procedure makes interpretation time-consuming, resource-hungry, and overall an expensive process. Automatic detection offers the potential to improve the quality of patient care by shortening the time to diagnosis, managing big data and optimizing the allocation of human resources towards precision medicine. Here, we present MindReader, a novel unsupervised machine-learning method comprised of the interplay between an autoencoder network, a hidden Markov model (HMM), and a generative component: after dividing the signal into overlapping frames and performing a fast Fourier transform, MindReader trains an autoencoder neural network for dimensionality reduction and compact representation of different frequency patterns for each frame. Next, we processed the temporal patterns using a HMM, while a third and generative component hypothesized and characterized the different phases that were then fed back to the HMM. MindReader then automatically generates labels that the physician can interpret as pathological and non-pathological phases, thus effectively reducing the search space for trained personnel. We evaluated MindReader's predictive performance on 686 recordings, encompassing more than 980 h from the publicly available Physionet database. Compared to manual annotations, MindReader identified 197 of 198 epileptic events (99.45%), and is, as such, a highly sensitive method, which is a prerequisite for clinical use.

Organ-specific metabolic pathways distinguish prediabetes, type 2 diabetes, and normal tissues.

Environmental and genetic factors cause defects in pancreatic islets driving type 2 diabetes (T2D) together with the progression of multi-tissue insulin resistance. Mass spectrometry proteomics on samples from five key metabolic tissues of a cross-sectional cohort of 43 multi-organ donors provides deep coverage of their proteomes. Enrichment analysis of Gene Ontology terms provides a tissue-specific map of altered biological processes across healthy, prediabetes (PD), and T2D subjects. We find widespread alterations in several relevant biological pathways, including increase in hemostasis in pancreatic islets of PD, increase in the complement cascade in liver and pancreatic islets of PD, and elevation in cholesterol biosynthesis in liver of T2D. Our findings point to inflammatory, immune, and vascular alterations in pancreatic islets in PD that are hypotheses to be tested for potential contributions to hormonal perturbations such as impaired insulin and increased glucagon production. This multi-tissue proteomic map suggests tissue-specific metabolic dysregulations in T2D.

Interpretable machine learning identifies paediatric Systemic Lupus Erythematosus subtypes based on gene expression data.

Transcriptomic analyses are commonly used to identify differentially expressed genes between patients and controls, or within individuals across disease courses. These methods, whilst effective, cannot encompass the combinatorial effects of genes driving disease. We applied rule-based machine learning (RBML) models and rule networks (RN) to an existing paediatric Systemic Lupus Erythematosus (SLE) blood expression dataset, with the goal of developing gene networks to separate low and high disease activity (DA1 and DA3). The resultant model had an 81% accuracy to distinguish between DA1 and DA3, with unsupervised hierarchical clustering revealing additional subgroups indicative of the immune axis involved or state of disease flare. These subgroups correlated with clinical variables, suggesting that the gene sets identified may further the understanding of gene networks that act in concert to drive disease progression. This included roles for genes (i) induced by interferons (IFI35 and OTOF), (ii) key to SLE cell types (KLRB1 encoding CD161), or (iii) with roles in autophagy and NF-κB pathway responses (CKAP4). As demonstrated here, RBML approaches have the potential to reveal novel gene patterns from within a heterogeneous disease, facilitating patient clinical and therapeutic stratification.

Machine Learning-Based Analysis of Glioma Grades Reveals Co-Enrichment.

Gliomas develop and grow in the brain and central nervous system. Examining glioma grading processes is valuable for improving therapeutic challenges. One of the most extensive repositories storing transcriptomics data for gliomas is The Cancer Genome Atlas (TCGA). However, such big cohorts should be processed with caution and evaluated thoroughly as they can contain batch and other effects. Furthermore, biological mechanisms of cancer contain interactions among biomarkers. Thus, we applied an interpretable machine learning approach to discover such relationships. This type of transparent learning provides not only good predictability, but also reveals co-predictive mechanisms among features. In this study, we corrected the strong and confounded batch effect in the TCGA glioma data. We further used the corrected datasets to perform comprehensive machine learning analysis applied on single-sample gene set enrichment scores using collections from the Molecular Signature Database. Furthermore, using rule-based classifiers, we displayed networks of co-enrichment related to glioma grades. Moreover, we validated our results using the external glioma cohorts. We believe that utilizing corrected glioma cohorts from TCGA may improve the application and validation of any future studies. Finally, the co-enrichment and survival analysis provided detailed explanations for glioma progression and consequently, it should support the targeted treatment.

Transcriptomic analysis reveals proinflammatory signatures associated with acute myeloid leukemia progression.

Numerous studies have been performed over the last decade to exploit the complexity of genomic and transcriptomic lesions driving the initiation of acute myeloid leukemia (AML). These studies have helped improve risk classification and treatment options. Detailed molecular characterization of longitudinal AML samples is sparse, however; meanwhile, relapse and therapy resistance represent the main challenges in AML care. To this end, we performed transcriptome-wide RNA sequencing of longitudinal diagnosis, relapse, and/or primary resistant samples from 47 adult and 23 pediatric AML patients with known mutational background. Gene expression analysis revealed the association of short event-free survival with overexpression of GLI2 and IL1R1, as well as downregulation of ST18. Moreover, CR1 downregulation and DPEP1 upregulation were associated with AML relapse both in adults and children. Finally, machine learning-based and network-based analysis identified overexpressed CD6 and downregulated INSR as highly copredictive genes depicting important relapse-associated characteristics among adult patients with AML. Our findings highlight the importance of a tumor-promoting inflammatory environment in leukemia progression, as indicated by several of the herein identified differentially expressed genes. Together, this knowledge provides the foundation for novel personalized drug targets and has the potential to maximize the benefit of current treatments to improve cure rates in AML.

MetaFetcheR: An R Package for Complete Mapping of Small-Compound Data.

Small-compound databases contain a large amount of information for metabolites and metabolic pathways. However, the plethora of such databases and the redundancy of their information lead to major issues with analysis and standardization. A lack of preventive establishment of means of data access at the infant stages of a project might lead to mislabelled compounds, reduced statistical power, and large delays in delivery of results. We developed MetaFetcheR, an open-source R package that links metabolite data from several small-compound databases, resolves inconsistencies, and covers a variety of use-cases of data fetching. We showed that the performance of MetaFetcheR was superior to existing approaches and databases by benchmarking the performance of the algorithm in three independent case studies based on two published datasets.

Interleukin-15 response signature predicts RhCMV/SIV vaccine efficacy.

Simian immunodeficiency virus (SIV) challenge of rhesus macaques (RMs) vaccinated with strain 68-1 Rhesus Cytomegalovirus (RhCMV) vectors expressing SIV proteins (RhCMV/SIV) results in a binary outcome: stringent control and subsequent clearance of highly pathogenic SIV in ~55% of vaccinated RMs with no protection in the remaining 45%. Although previous work indicates that unconventionally restricted, SIV-specific, effector-memory (EM)-biased CD8+ T cell responses are necessary for efficacy, the magnitude of these responses does not predict efficacy, and the basis of protection vs. non-protection in 68-1 RhCMV/SIV vector-vaccinated RMs has not been elucidated. Here, we report that 68-1 RhCMV/SIV vector administration strikingly alters the whole blood transcriptome of vaccinated RMs, with the sustained induction of specific immune-related pathways, including immune cell, toll-like receptor (TLR), inflammasome/cell death, and interleukin-15 (IL-15) signaling, significantly correlating with subsequent vaccine efficacy. Treatment of a separate RM cohort with IL-15 confirmed the central involvement of this cytokine in the protection signature, linking the major innate and adaptive immune gene expression networks that correlate with RhCMV/SIV vaccine efficacy. This change-from-baseline IL-15 response signature was also demonstrated to significantly correlate with vaccine efficacy in an independent validation cohort of vaccinated and challenged RMs. The differential IL-15 gene set response to vaccination strongly correlated with the pre-vaccination activity of this pathway, with reduced baseline expression of IL-15 response genes significantly correlating with higher vaccine-induced induction of IL-15 signaling and subsequent vaccine protection, suggesting that a robust de novo vaccine-induced IL-15 signaling response is needed to program vaccine efficacy. Thus, the RhCMV/SIV vaccine imparts a coordinated and persistent induction of innate and adaptive immune pathways featuring IL-15, a known regulator of CD8+ T cell function, that support the ability of vaccine-elicited unconventionally restricted CD8+ T cells to mediate protection against SIV challenge.

Functional annotation of noncoding mutations in cancer.

In a cancer genome, the noncoding sequence contains the vast majority of somatic mutations. While very few are expected to be cancer drivers, those affecting regulatory elements have the potential to have downstream effects on gene regulation that may contribute to cancer progression. To prioritize regulatory mutations, we screened somatic mutations in the Pan-Cancer Analysis of Whole Genomes cohort of 2,515 cancer genomes on individual bases to assess their potential regulatory roles in their respective cancer types. We found a highly significant enrichment of regulatory mutations associated with the deamination signature overlapping a CpG site in the CCAAT/Enhancer Binding Protein ß recognition sites in many cancer types. Overall, 5,749 mutated regulatory elements were identified in 1,844 tumor samples from 39 cohorts containing 11,962 candidate regulatory mutations. Our analysis indicated 20 or more regulatory mutations in 5.5% of the samples, and an overall average of six per tumor. Several recurrent elements were identified, and major cancer-related pathways were significantly enriched for genes nearby the mutated regulatory elements. Our results provide a detailed view of the role of regulatory elements in cancer genomes.

Multifaceted regulation of hepatic lipid metabolism by YY1.

Recent studies suggested that dysregulated YY1 plays a pivotal role in many liver diseases. To obtain a detailed view of genes and pathways regulated by YY1 in the liver, we carried out RNA sequencing in HepG2 cells after YY1 knockdown. A rigid set of 2,081 differentially expressed genes was identified by comparing the YY1-knockdown samples (n = 8) with the control samples (n = 14). YY1 knockdown significantly decreased the expression of several key transcription factors and their coactivators in lipid metabolism. This is illustrated by YY1 regulating PPARA expression through binding to its promoter and enhancer regions. Our study further suggest that down-regulation of the key transcription factors together with YY1 knockdown significantly decreased the cooperation between YY1 and these transcription factors at various regulatory regions, which are important in regulating the expression of genes in hepatic lipid metabolism. This was supported by the finding that the expression of SCD and ELOVL6, encoding key enzymes in lipogenesis, were regulated by the cooperation between YY1 and PPARA/RXRA complex over their promoters.

Mapping chromatin accessibility and active regulatory elements reveals pathological mechanisms in human gliomas.

Chromatin structure and accessibility, and combinatorial binding of transcription factors to regulatory elements in genomic DNA control transcription. Genetic variations in genes encoding histones, epigenetics-related enzymes or modifiers affect chromatin structure/dynamics and result in alterations in gene expression contributing to cancer development or progression. Gliomas are brain tumors frequently associated with epigenetics-related gene deregulation. We perform whole-genome mapping of chromatin accessibility, histone modifications, DNA methylation patterns and transcriptome analysis simultaneously in multiple tumor samples to unravel epigenetic dysfunctions driving gliomagenesis. Based on the results of the integrative analysis of the acquired profiles, we create an atlas of active enhancers and promoters in benign and malignant gliomas. We explore these elements and intersect with Hi-C data to uncover molecular mechanisms instructing gene expression in gliomas.

Interpretable Machine Learning Reveals Dissimilarities Between Subtypes of Autism Spectrum Disorder.

Autism spectrum disorder (ASD) is a heterogeneous neuropsychiatric disorder with a complex genetic background. Analysis of altered molecular processes in ASD patients requires linear and nonlinear methods that provide interpretable solutions. Interpretable machine learning provides legible models that allow explaining biological mechanisms and support analysis of clinical subgroups. In this work, we investigated several case-control studies of gene expression measurements of ASD individuals. We constructed a rule-based learning model from three independent datasets that we further visualized as a nonlinear gene-gene co-predictive network. To find dissimilarities between ASD subtypes, we scrutinized a topological structure of the network and estimated a centrality distance. Our analysis revealed that autism is the most severe subtype of ASD, while pervasive developmental disorder-not otherwise specified and Asperger syndrome are closely related and milder ASD subtypes. Furthermore, we analyzed the most important ASD-related features that were described in terms of gene co-predictors. Among others, we found a strong co-predictive mechanism between EMC4 and TMEM30A, which may suggest a co-regulation between these genes. The present study demonstrates the potential of applying interpretable machine learning in bioinformatics analyses. Although the proposed methodology was designed for transcriptomics data, it can be applied to other omics disciplines.

R.ROSETTA: an interpretable machine learning framework.

BACKGROUND: Machine learning involves strategies and algorithms that may assist bioinformatics analyses in terms of data mining and knowledge discovery. In several applications, viz. in Life Sciences, it is often more important to understand how a prediction was obtained rather than knowing what prediction was made. To this end so-called interpretable machine learning has been recently advocated. In this study, we implemented an interpretable machine learning package based on the rough set theory. An important aim of our work was provision of statistical properties of the models and their components. RESULTS: We present the R.ROSETTA package, which is an R wrapper of ROSETTA framework. The original ROSETTA functions have been improved and adapted to the R programming environment. The package allows for building and analyzing non-linear interpretable machine learning models. R.ROSETTA gathers combinatorial statistics via rule-based modelling for accessible and transparent results, well-suited for adoption within the greater scientific community. The package also provides statistics and visualization tools that facilitate minimization of analysis bias and noise. The R.ROSETTA package is freely available at https://github.com/komorowskilab/R.ROSETTA . To illustrate the usage of the package, we applied it to a transcriptome dataset from an autism case-control study. Our tool provided hypotheses for potential co-predictive mechanisms among features that discerned phenotype classes. These co-predictors represented neurodevelopmental and autism-related genes. CONCLUSIONS: R.ROSETTA provides new insights for interpretable machine learning analyses and knowledge-based systems. We demonstrated that our package facilitated detection of dependencies for autism-related genes. Although the sample application of R.ROSETTA illustrates transcriptome data analysis, the package can be used to analyze any data organized in decision tables.

Genomic characterization of relapsed acute myeloid leukemia reveals novel putative therapeutic targets.

Relapse is the leading cause of death of adult and pediatric patients with acute myeloid leukemia (AML). Numerous studies have helped to elucidate the complex mutational landscape at diagnosis of AML, leading to improved risk stratification and new therapeutic options. However, multi-whole-genome studies of adult and pediatric AML at relapse are necessary for further advances. To this end, we performed whole-genome and whole-exome sequencing analyses of longitudinal diagnosis, relapse, and/or primary resistant specimens from 48 adult and 25 pediatric patients with AML. We identified mutations recurrently gained at relapse in ARID1A and CSF1R, both of which represent potentially actionable therapeutic alternatives. Further, we report specific differences in the mutational spectrum between adult vs pediatric relapsed AML, with MGA and H3F3A p.Lys28Met mutations recurrently found at relapse in adults, whereas internal tandem duplications in UBTF were identified solely in children. Finally, our study revealed recurrent mutations in IKZF1, KANSL1, and NIPBL at relapse. All of the mentioned genes have either never been reported at diagnosis in de novo AML or have been reported at low frequency, suggesting important roles for these alterations predominantly in disease progression and/or resistance to therapy. Our findings shed further light on the complexity of relapsed AML and identified previously unappreciated alterations that may lead to improved outcomes through personalized medicine.

Nucleolar rDNA folds into condensed foci with a specific combination of epigenetic marks.

Arabidopsis thaliana 45S ribosomal genes (rDNA) are located in tandem arrays called nucleolus organizing regions on the termini of chromosomes 2 and 4 (NOR2 and NOR4) and encode rRNA, a crucial structural element of the ribosome. The current model of rDNA organization suggests that inactive rRNA genes accumulate in the condensed chromocenters in the nucleus and at the nucleolar periphery, while the nucleolus delineates active genes. We challenge the perspective that all intranucleolar rDNA is active by showing that a subset of nucleolar rDNA assembles into condensed foci marked by H3.1 and H3.3 histones that also contain the repressive H3K9me2 histone mark. By using plant lines containing a low number of rDNA copies, we further found that the condensed foci relate to the folding of rDNA, which appears to be a common mechanism of rDNA regulation inside the nucleolus. The H3K9me2 histone mark found in condensed foci represents a typical modification of bulk inactive rDNA, as we show by genome-wide approaches, similar to the H2A.W histone variant. The euchromatin histone marks H3K27me3 and H3K4me3, in contrast, do not colocalize with nucleolar foci and their overall levels in the nucleolus are very low. We further demonstrate that the rDNA promoter is an important regulatory region of the rDNA, where the distribution of histone variants and histone modifications are modulated in response to rDNA activity.

Integration of whole-body [18F]FDG PET/MRI with non-targeted metabolomics can provide new insights on tissue-specific insulin resistance in type 2 diabetes.

Alteration of various metabolites has been linked to type 2 diabetes (T2D) and insulin resistance. However, identifying significant associations between metabolites and tissue-specific phenotypes requires a multi-omics approach. In a cohort of 42 subjects with different levels of glucose tolerance (normal, prediabetes and T2D) matched for age and body mass index, we calculated associations between parameters of whole-body positron emission tomography (PET)/magnetic resonance imaging (MRI) during hyperinsulinemic euglycemic clamp and non-targeted metabolomics profiling for subcutaneous adipose tissue (SAT) and plasma. Plasma metabolomics profiling revealed that hepatic fat content was positively associated with tyrosine, and negatively associated with lysoPC(P-16:0). Visceral adipose tissue (VAT) and SAT insulin sensitivity (Ki), were positively associated with several lysophospholipids, while the opposite applied to branched-chain amino acids. The adipose tissue metabolomics revealed a positive association between non-esterified fatty acids and, VAT and liver Ki. Bile acids and carnitines in adipose tissue were inversely associated with VAT Ki. Furthermore, we detected several metabolites that were significantly higher in T2D than normal/prediabetes. In this study we present novel associations between several metabolites from SAT and plasma with the fat fraction, volume and insulin sensitivity of various tissues throughout the body, demonstrating the benefit of an integrative multi-omics approach.

A Multi-Omics Approach to Liver Diseases: Integration of Single Nuclei Transcriptomics with Proteomics and HiCap Bulk Data in Human Liver.

The liver is the largest solid organ and a primary metabolic hub. In recent years, intact cell nuclei were used to perform single-nuclei RNA-seq (snRNA-seq) for tissues difficult to dissociate and for flash-frozen archived tissue samples to discover unknown and rare cell subpopulations. In this study, we performed snRNA-seq of a liver sample to identify subpopulations of cells based on nuclear transcriptomics. In 4282 single nuclei, we detected, on average, 1377 active genes and we identified seven major cell types. We integrated data from 94,286 distal interactions (p < 0.05) for 7682 promoters from a targeted chromosome conformation capture technique (HiCap) and mass spectrometry proteomics for the same liver sample. We observed a reasonable correlation between proteomics and in silico bulk snRNA-seq (r = 0.47) using tissue-independent gene-specific protein abundancy estimation factors. We specifically looked at genes of medical importance. The DPYD gene is involved in the pharmacogenetics of fluoropyrimidine toxicity and some of its variants are analyzed for clinical purposes. We identified a new putative polymorphic regulatory element, which may contribute to variation in toxicity. Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer and we investigated all known risk genes. We identified a complex regulatory landscape for the SLC2A2 gene with 16 candidate enhancers. Three of them harbor somatic motif breaking and other mutations in HCC in the Pan Cancer Analysis of Whole Genomes dataset and are candidates to contribute to malignancy. Our results highlight the potential of a multi-omics approach in the study of human diseases.

Analyses of non-coding somatic drivers in 2,658 cancer whole genomes.

The discovery of drivers of cancer has traditionally focused on protein-coding genes1-4. Here we present analyses of driver point mutations and structural variants in non-coding regions across 2,658 genomes from the Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium5 of the International Cancer Genome Consortium (ICGC) and The Cancer Genome Atlas (TCGA). For point mutations, we developed a statistically rigorous strategy for combining significance levels from multiple methods of driver discovery that overcomes the limitations of individual methods. For structural variants, we present two methods of driver discovery, and identify regions that are significantly affected by recurrent breakpoints and recurrent somatic juxtapositions. Our analyses confirm previously reported drivers6,7, raise doubts about others and identify novel candidates, including point mutations in the 5' region of TP53, in the 3' untranslated regions of NFKBIZ and TOB1, focal deletions in BRD4 and rearrangements in the loci of AKR1C genes. We show that although point mutations and structural variants that drive cancer are less frequent in non-coding genes and regulatory sequences than in protein-coding genes, additional examples of these drivers will be found as more cancer genomes become available.

Author Correction: Macrophage-associated wound healing contributes to African green monkey SIV pathogenesis control.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.