Lifestyle integrated functional exercise for people with interstitial lung disease (iLiFE): A mixed-methods feasibility study.

BACKGROUND: People with interstitial lung disease (ILD) present low levels of physical activity (PA) and spend most of their time at home, especially in advanced stages of the disease. The Lifestyle Integrated Functional Exercise for people with ILD (iLiFE) embedding PA in patients' daily routines was developed and implemented. OBJECTIVES: This study aimed to explore the feasibility of iLiFE. METHODS: A pre/post mixed-methods feasibility study was conducted. Feasibility of iLiFE was determined by participant recruitment/retention, adherence, feasibility of outcome measures and adverse events. Measures of PA, sedentary behaviour, balance, muscle strength, functional performance/capacity, exercise capacity, impact of the disease, symptoms (i.e., dyspnoea, anxiety, depression, fatigue and cough) and health-related quality of life were collected at baseline and post-intervention (12-weeks). Semi-structured interviews with participants were conducted in-person immediately after iLiFE. Interviews were audio-recorded, transcribed and analysed by deductive thematic analysis. RESULTS: Ten participants (5♀, 77±3y; FVCpp 77.1 ± 4.4, DLCOpp 42.4 ± 6.6) were included, but only nine completed the study. Recruitment was challenging (30%) and retention high (90%). iLiFE was feasible, with excellent adherence (84.4%) and no adverse events. Missing data were associated with one dropout and non-compliance with the accelerometer (n = 1). Participants reported that iLiFE contributed to (re)gain control in their daily life, namely through improving their well-being, functional status and motivation. Weather, symptoms, physical impairments and lack of motivation were identified as threats to keep an active lifestyle. CONCLUSIONS: iLiFE seems to be feasible, safe and meaningful for people with ILD. A randomised controlled trial is needed to strengthen these promising findings.

A systematic evaluation of deep learning methods for the prediction of drug synergy in cancer.

One of the main obstacles to the successful treatment of cancer is the phenomenon of drug resistance. A common strategy to overcome resistance is the use of combination therapies. However, the space of possibilities is huge and efficient search strategies are required. Machine Learning (ML) can be a useful tool for the discovery of novel, clinically relevant anti-cancer drug combinations. In particular, deep learning (DL) has become a popular choice for modeling drug combination effects. Here, we set out to examine the impact of different methodological choices on the performance of multimodal DL-based drug synergy prediction methods, including the use of different input data types, preprocessing steps and model architectures. Focusing on the NCI ALMANAC dataset, we found that feature selection based on prior biological knowledge has a positive impact-limiting gene expression data to cancer or drug response-specific genes improved performance. Drug features appeared to be more predictive of drug response, with a 41% increase in coefficient of determination (R2) and 26% increase in Spearman correlation relative to a baseline model that used only cell line and drug identifiers. Molecular fingerprint-based drug representations performed slightly better than learned representations-ECFP4 fingerprints increased R2 by 5.3% and Spearman correlation by 2.8% w.r.t the best learned representations. In general, fully connected feature-encoding subnetworks outperformed other architectures. DL outperformed other ML methods by more than 35% (R2) and 14% (Spearman). Additionally, an ensemble combining the top DL and ML models improved performance by about 6.5% (R2) and 4% (Spearman). Using a state-of-the-art interpretability method, we showed that DL models can learn to associate drug and cell line features with drug response in a biologically meaningful way. The strategies explored in this study will help to improve the development of computational methods for the rational design of effective drug combinations for cancer therapy.

The landscape of expression and alternative splicing variation across human traits.

Understanding the consequences of individual transcriptome variation is fundamental to deciphering human biology and disease. We implement a statistical framework to quantify the contributions of 21 individual traits as drivers of gene expression and alternative splicing variation across 46 human tissues and 781 individuals from the Genotype-Tissue Expression project. We demonstrate that ancestry, sex, age, and BMI make additive and tissue-specific contributions to expression variability, whereas interactions are rare. Variation in splicing is dominated by ancestry and is under genetic control in most tissues, with ribosomal proteins showing a strong enrichment of tissue-shared splicing events. Our analyses reveal a systemic contribution of types 1 and 2 diabetes to tissue transcriptome variation with the strongest signal in the nerve, where histopathology image analysis identifies novel genes related to diabetic neuropathy. Our multi-tissue and multi-trait approach provides an extensive characterization of the main drivers of human transcriptome variation in health and disease.

Scalable transcriptomics analysis with Dask: applications in data science and machine learning.

BACKGROUND: Gene expression studies are an important tool in biological and biomedical research. The signal carried in expression profiles helps derive signatures for the prediction, diagnosis and prognosis of different diseases. Data science and specifically machine learning have many applications in gene expression analysis. However, as the dimensionality of genomics datasets grows, scalable solutions become necessary. METHODS: In this paper we review the main steps and bottlenecks in machine learning pipelines, as well as the main concepts behind scalable data science including those of concurrent and parallel programming. We discuss the benefits of the Dask framework and how it can be integrated with the Python scientific environment to perform data analysis in computational biology and bioinformatics. RESULTS: This review illustrates the role of Dask for boosting data science applications in different case studies. Detailed documentation and code on these procedures is made available at https://github.com/martaccmoreno/gexp-ml-dask . CONCLUSION: By showing when and how Dask can be used in transcriptomics analysis, this review will serve as an entry point to help genomic data scientists develop more scalable data analysis procedures.

Deep learning for drug response prediction in cancer.

Predicting the sensitivity of tumors to specific anti-cancer treatments is a challenge of paramount importance for precision medicine. Machine learning(ML) algorithms can be trained on high-throughput screening data to develop models that are able to predict the response of cancer cell lines and patients to novel drugs or drug combinations. Deep learning (DL) refers to a distinct class of ML algorithms that have achieved top-level performance in a variety of fields, including drug discovery. These types of models have unique characteristics that may make them more suitable for the complex task of modeling drug response based on both biological and chemical data, but the application of DL to drug response prediction has been unexplored until very recently. The few studies that have been published have shown promising results, and the use of DL for drug response prediction is beginning to attract greater interest from researchers in the field. In this article, we critically review recently published studies that have employed DL methods to predict drug response in cancer cell lines. We also provide a brief description of DL and the main types of architectures that have been used in these studies. Additionally, we present a selection of publicly available drug screening data resources that can be used to develop drug response prediction models. Finally, we also address the limitations of these approaches and provide a discussion on possible paths for further improvement. Contact:  mrocha@di.uminho.pt.

Gender Differential Transcriptome in Gastric and Thyroid Cancers.

Cancer has an important and considerable gender differential susceptibility confirmed by several epidemiological studies. Gastric (GC) and thyroid cancer (TC) are examples of malignancies with a higher incidence in males and females, respectively. Beyond environmental predisposing factors, it is expected that gender-specific gene deregulation contributes to this differential incidence. We performed a detailed characterization of the transcriptomic differences between genders in normal and tumor tissues from stomach and thyroid using Genotype-Tissue Expression (GTEx) and The Cancer Genome Atlas (TCGA) data. We found hundreds of sex-biased genes (SBGs). Most of the SBGs shared by normal and tumor belong to sexual chromosomes, while the normal and tumor-specific tend to be found in the autosomes. Expression of several cancer-associated genes is also found to differ between sexes in both types of tissue. Thousands of differentially expressed genes (DEGs) between paired tumor-normal tissues were identified in GC and TC. For both cancers, in the most susceptible gender, the DEGs were mostly under-expressed in the tumor tissue, with an enrichment for tumor-suppressor genes (TSGs). Moreover, we found gene networks preferentially associated to males in GC and to females in TC and correlated with cancer histological subtypes. Our results shed light on the molecular differences and commonalities between genders and provide novel insights in the differential risk underlying these cancers.

Cosmology and fundamental physics with the Euclid satellite.

Euclid is a European Space Agency medium-class mission selected for launch in 2020 within the cosmic vision 2015-2025 program. The main goal of Euclid is to understand the origin of the accelerated expansion of the universe. Euclid will explore the expansion history of the universe and the evolution of cosmic structures by measuring shapes and red-shifts of galaxies as well as the distribution of clusters of galaxies over a large fraction of the sky. Although the main driver for Euclid is the nature of dark energy, Euclid science covers a vast range of topics, from cosmology to galaxy evolution to planetary research. In this review we focus on cosmology and fundamental physics, with a strong emphasis on science beyond the current standard models. We discuss five broad topics: dark energy and modified gravity, dark matter, initial conditions, basic assumptions and questions of methodology in the data analysis. This review has been planned and carried out within Euclid's Theory Working Group and is meant to provide a guide to the scientific themes that will underlie the activity of the group during the preparation of the Euclid mission.

The effects of death and post-mortem cold ischemia on human tissue transcriptomes.

Post-mortem tissues samples are a key resource for investigating patterns of gene expression. However, the processes triggered by death and the post-mortem interval (PMI) can significantly alter physiologically normal RNA levels. We investigate the impact of PMI on gene expression using data from multiple tissues of post-mortem donors obtained from the GTEx project. We find that many genes change expression over relatively short PMIs in a tissue-specific manner, but this potentially confounding effect in a biological analysis can be minimized by taking into account appropriate covariates. By comparing ante- and post-mortem blood samples, we identify the cascade of transcriptional events triggered by death of the organism. These events do not appear to simply reflect stochastic variation resulting from mRNA degradation, but active and ongoing regulation of transcription. Finally, we develop a model to predict the time since death from the analysis of the transcriptome of a few readily accessible tissues.

Time-dependent genetic effects on gene expression implicate aging processes.

Gene expression is dependent on genetic and environmental factors. In the last decade, a large body of research has significantly improved our understanding of the genetic architecture of gene expression. However, it remains unclear whether genetic effects on gene expression remain stable over time. Here, we show, using longitudinal whole-blood gene expression data from a twin cohort, that the genetic architecture of a subset of genes is unstable over time. In addition, we identified 2213 genes differentially expressed across time points that we linked with aging within and across studies. Interestingly, we discovered that most differentially expressed genes were affected by a subset of 77 putative causal genes. Finally, we observed that putative causal genes and down-regulated genes were affected by a loss of genetic control between time points. Taken together, our data suggest that instability in the genetic architecture of a subset of genes could lead to widespread effects on the transcriptome with an aging signature.

Sequence variation between 462 human individuals fine-tunes functional sites of RNA processing.

Recent advances in the cost-efficiency of sequencing technologies enabled the combined DNA- and RNA-sequencing of human individuals at the population-scale, making genome-wide investigations of the inter-individual genetic impact on gene expression viable. Employing mRNA-sequencing data from the Geuvadis Project and genome sequencing data from the 1000 Genomes Project we show that the computational analysis of DNA sequences around splice sites and poly-A signals is able to explain several observations in the phenotype data. In contrast to widespread assessments of statistically significant associations between DNA polymorphisms and quantitative traits, we developed a computational tool to pinpoint the molecular mechanisms by which genetic markers drive variation in RNA-processing, cataloguing and classifying alleles that change the affinity of core RNA elements to their recognizing factors. The in silico models we employ further suggest RNA editing can moonlight as a splicing-modulator, albeit less frequently than genomic sequence diversity. Beyond existing annotations, we demonstrate that the ultra-high resolution of RNA-Seq combined from 462 individuals also provides evidence for thousands of bona fide novel elements of RNA processing-alternative splice sites, introns, and cleavage sites-which are often rare and lowly expressed but in other characteristics similar to their annotated counterparts.

Spatiotemporal Control of Forkhead Binding to DNA Regulates the Meiotic Gene Expression Program.

Meiosis is a differentiated program of the cell cycle that is characterized by high levels of recombination followed by two nuclear divisions. In fission yeast, the genetic program during meiosis is regulated at multiple levels, including transcription, mRNA stabilization, and splicing. Mei4 is a forkhead transcription factor that controls the expression of mid-meiotic genes. Here, we describe that Fkh2, another forkhead transcription factor that is essential for mitotic cell-cycle progression, also plays a pivotal role in the control of meiosis. Fkh2 binding preexists in most Mei4-dependent genes, inhibiting their expression. During meiosis, Fkh2 is phosphorylated in a CDK/Cig2-dependent manner, decreasing its affinity for DNA, which creates a window of opportunity for Mei4 binding to its target genes. We propose that Fkh2 serves as a placeholder until the later appearance of Mei4 with a higher affinity for DNA that induces the expression of a subset of meiotic genes.

Molecular signatures of plastic phenotypes in two eusocial insect species with simple societies.

Phenotypic plasticity is important in adaptation and shapes the evolution of organisms. However, we understand little about what aspects of the genome are important in facilitating plasticity. Eusocial insect societies produce plastic phenotypes from the same genome, as reproductives (queens) and nonreproductives (workers). The greatest plasticity is found in the simple eusocial insect societies in which individuals retain the ability to switch between reproductive and nonreproductive phenotypes as adults. We lack comprehensive data on the molecular basis of plastic phenotypes. Here, we sequenced genomes, microRNAs (miRNAs), and multiple transcriptomes and methylomes from individual brains in a wasp (Polistes canadensis) and an ant (Dinoponera quadriceps) that live in simple eusocial societies. In both species, we found few differences between phenotypes at the transcriptional level, with little functional specialization, and no evidence that phenotype-specific gene expression is driven by DNA methylation or miRNAs. Instead, phenotypic differentiation was defined more subtly by nonrandom transcriptional network organization, with roles in these networks for both conserved and taxon-restricted genes. The general lack of highly methylated regions or methylome patterning in both species may be an important mechanism for achieving plasticity among phenotypes during adulthood. These findings define previously unidentified hypotheses on the genomic processes that facilitate plasticity and suggest that the molecular hallmarks of social behavior are likely to differ with the level of social complexity.

Human genomics. The human transcriptome across tissues and individuals.

Transcriptional regulation and posttranscriptional processing underlie many cellular and organismal phenotypes. We used RNA sequence data generated by Genotype-Tissue Expression (GTEx) project to investigate the patterns of transcriptome variation across individuals and tissues. Tissues exhibit characteristic transcriptional signatures that show stability in postmortem samples. These signatures are dominated by a relatively small number of genes­which is most clearly seen in blood­though few are exclusive to a particular tissue and vary more across tissues than individuals. Genes exhibiting high interindividual expression variation include disease candidates associated with sex, ethnicity, and age. Primary transcription is the major driver of cellular specificity, with splicing playing mostly a complementary role; except for the brain, which exhibits a more divergent splicing program. Variation in splicing, despite its stochasticity, may play in contrast a comparatively greater role in defining individual phenotypes.

Human genomics. Effect of predicted protein-truncating genetic variants on the human transcriptome.

Accurate prediction of the functional effect of genetic variation is critical for clinical genome interpretation. We systematically characterized the transcriptome effects of protein-truncating variants, a class of variants expected to have profound effects on gene function, using data from the Genotype-Tissue Expression (GTEx) and Geuvadis projects. We quantitated tissue-specific and positional effects on nonsense-mediated transcript decay and present an improved predictive model for this decay. We directly measured the effect of variants both proximal and distal to splice junctions. Furthermore, we found that robustness to heterozygous gene inactivation is not due to dosage compensation. Our results illustrate the value of transcriptome data in the functional interpretation of genetic variants.

Short term exposure of beta cells to low concentrations of interleukin-1ß improves insulin secretion through focal adhesion and actin remodeling and regulation of gene expression.

Type 2 diabetes involves defective insulin secretion with islet inflammation governed in part by IL-1ß. Prolonged exposure of islets to high concentrations of IL-1ß (>24 h, 20 ng/ml) impairs beta cell function and survival. Conversely, exposure to lower concentrations of IL-1ß for >24 h improves these same parameters. The impact on insulin secretion of shorter exposure times to IL-1ß and the underlying molecular mechanisms are poorly understood and were the focus of this study. Treatment of rat primary beta cells, as well as rat or human whole islets, with 0.1 ng/ml IL-1ß for 2 h increased glucose-stimulated (but not basal) insulin secretion, whereas 20 ng/ml was without effect. Similar differential effects of IL-1ß depending on concentration were observed after 15 min of KCl stimulation but were prevented by diazoxide. Studies on sorted rat beta cells indicated that the enhancement of stimulated secretion by 0.1 ng/ml IL-1ß was mediated by the NF-κB pathway and c-JUN/JNK pathway acting in parallel to elicit focal adhesion remodeling and the phosphorylation of paxillin independently of upstream regulation by focal adhesion kinase. Because the beneficial effect of IL-1ß was dependent in part upon transcription, gene expression was analyzed by RNAseq. There were 18 genes regulated uniquely by 0.1 but not 20 ng/ml IL-1ß, which are mostly involved in transcription and apoptosis. These results indicate that 2 h of exposure of beta cells to a low but not a high concentration of IL-1ß enhances glucose-stimulated insulin secretion through focal adhesion and actin remodeling, as well as modulation of gene expression.

Biased allelic expression in human primary fibroblast single cells.

The study of gene expression in mammalian single cells via genomic technologies now provides the possibility to investigate the patterns of allelic gene expression. We used single-cell RNA sequencing to detect the allele-specific mRNA level in 203 single human primary fibroblasts over 133,633 unique heterozygous single-nucleotide variants (hetSNVs). We observed that at the snapshot of analyses, each cell contained mostly transcripts from one allele from the majority of genes; indeed, 76.4% of the hetSNVs displayed stochastic monoallelic expression in single cells. Remarkably, adjacent hetSNVs exhibited a haplotype-consistent allelic ratio; in contrast, distant sites located in two different genes were independent of the haplotype structure. Moreover, the allele-specific expression in single cells correlated with the abundance of the cellular transcript. We observed that genes expressing both alleles in the majority of the single cells at a given time point were rare and enriched with highly expressed genes. The relative abundance of each allele in a cell was controlled by some regulatory mechanisms given that we observed related single-cell allelic profiles according to genes. Overall, these results have direct implications in cellular phenotypic variability.

Putative cis-regulatory drivers in colorectal cancer.

The cis-regulatory effects responsible for cancer development have not been as extensively studied as the perturbations of the protein coding genome in tumorigenesis. To better characterize colorectal cancer (CRC) development we conducted an RNA-sequencing experiment of 103 matched tumour and normal colon mucosa samples from Danish CRC patients, 90 of which were germline-genotyped. By investigating allele-specific expression (ASE) we show that the germline genotypes remain important determinants of allelic gene expression in tumours. Using the changes in ASE in matched pairs of samples we discover 71 genes with excess of somatic cis-regulatory effects in CRC, suggesting a cancer driver role. We correlate genotypes and gene expression to identify expression quantitative trait loci (eQTLs) and find 1,693 and 948 eQTLs in normal samples and tumours, respectively. We estimate that 36% of the tumour eQTLs are exclusive to CRC and show that this specificity is partially driven by increased expression of specific transcription factors and changes in methylation patterns. We show that tumour-specific eQTLs are more enriched for low CRC genome-wide association study (GWAS) P values than shared eQTLs, which suggests that some of the GWAS variants are tumour specific regulatory variants. Importantly, tumour-specific eQTL genes also accumulate more somatic mutations when compared to the shared eQTL genes, raising the possibility that they constitute germline-derived cancer regulatory drivers. Collectively the integration of genome and the transcriptome reveals a substantial number of putative somatic and germline cis-regulatory cancer changes that may have a role in tumorigenesis.

Tandem RNA chimeras contribute to transcriptome diversity in human population and are associated with intronic genetic variants.

Chimeric RNAs originating from two or more different genes are known to exist not only in cancer, but also in normal tissues, where they can play a role in human evolution. However, the exact mechanism of their formation is unknown. Here, we use RNA sequencing data from 462 healthy individuals representing 5 human populations to systematically identify and in depth characterize 81 RNA tandem chimeric transcripts, 13 of which are novel. We observe that 6 out of these 81 chimeras have been regarded as cancer-specific. Moreover, we show that a prevalence of long introns at the fusion breakpoint is associated with the chimeric transcripts formation. We also find that tandem RNA chimeras have lower abundances as compared to their partner genes. Finally, by combining our results with genomic data from the same individuals we uncover intronic genetic variants associated with the chimeric RNA formation. Taken together our findings provide an important insight into the chimeric transcripts formation and open new avenues of research into the role of intronic genetic variants in post-transcriptional processing events.

Identification of genetic variants associated with alternative splicing using sQTLseekeR.

Identification of genetic variants affecting splicing in RNA sequencing population studies is still in its infancy. Splicing phenotype is more complex than gene expression and ought to be treated as a multivariate phenotype to be recapitulated completely. Here we represent the splicing pattern of a gene as the distribution of the relative abundances of a gene's alternative transcript isoforms. We develop a statistical framework that uses a distance-based approach to compute the variability of splicing ratios across observations, and a non-parametric analogue to multivariate analysis of variance. We implement this approach in the R package sQTLseekeR and use it to analyze RNA-Seq data from the Geuvadis project in 465 individuals. We identify hundreds of single nucleotide polymorphisms (SNPs) as splicing QTLs (sQTLs), including some falling in genome-wide association study SNPs. By developing the appropriate metrics, we show that sQTLseekeR compares favorably with existing methods that rely on univariate approaches, predicting variants that behave as expected from mutations affecting splicing.