Showing papers by "Pompeu Fabra University published in 2018"

Identifying the Best Machine Learning Algorithms for Brain Tumor Segmentation, Progression Assessment, and Overall Survival Prediction in the BRATS Challenge

Abstract: Gliomas are the most common primary brain malignancies, with different degrees of aggressiveness, variable prognosis and various heterogeneous histologic sub-regions, i.e., peritumoral edematous/invaded tissue, necrotic core, active and non-enhancing core. This intrinsic heterogeneity is also portrayed in their radio-phenotype, as their sub-regions are depicted by varying intensity profiles disseminated across multi-parametric magnetic resonance imaging (mpMRI) scans, reflecting varying biological properties. Their heterogeneous shape, extent, and location are some of the factors that make these tumors difficult to resect, and in some cases inoperable. The amount of resected tumoris a factor also considered in longitudinal scans, when evaluating the apparent tumor for potential diagnosis of progression. Furthermore, there is mounting evidence that accurate segmentation of the various tumor sub-regions can offer the basis for quantitative image analysis towards prediction of patient overall survival. This study assesses thestate-of-the-art machine learning (ML) methods used for brain tumor image analysis in mpMRI scans, during the last seven instances of the International Brain Tumor Segmentation (BraTS) challenge, i.e., 2012-2018. Specifically, we focus on i) evaluating segmentations of the various glioma sub-regions in pre-operative mpMRI scans, ii) assessing potential tumor progression by virtue of longitudinal growth of tumor sub-regions, beyond use of the RECIST/RANO criteria, and iii) predicting the overall survival from pre-operative mpMRI scans of patients that underwent gross tota lresection. Finally, we investigate the challenge of identifying the best ML algorithms for each of these tasks, considering that apart from being diverse on each instance of the challenge, the multi-institutional mpMRI BraTS dataset has also been a continuously evolving/growing dataset.

Deep Learning Techniques for Automatic MRI Cardiac Multi-Structures Segmentation and Diagnosis: Is the Problem Solved?

Abstract: Delineation of the left ventricular cavity, myocardium, and right ventricle from cardiac magnetic resonance images (multi-slice 2-D cine MRI) is a common clinical task to establish diagnosis. The automation of the corresponding tasks has thus been the subject of intense research over the past decades. In this paper, we introduce the “Automatic Cardiac Diagnosis Challenge” dataset (ACDC), the largest publicly available and fully annotated dataset for the purpose of cardiac MRI (CMR) assessment. The dataset contains data from 150 multi-equipments CMRI recordings with reference measurements and classification from two medical experts. The overarching objective of this paper is to measure how far state-of-the-art deep learning methods can go at assessing CMRI, i.e., segmenting the myocardium and the two ventricles as well as classifying pathologies. In the wake of the 2017 MICCAI-ACDC challenge, we report results from deep learning methods provided by nine research groups for the segmentation task and four groups for the classification task. Results show that the best methods faithfully reproduce the expert analysis, leading to a mean value of 0.97 correlation score for the automatic extraction of clinical indices and an accuracy of 0.96 for automatic diagnosis. These results clearly open the door to highly accurate and fully automatic analysis of cardiac CMRI. We also identify scenarios for which deep learning methods are still failing. Both the dataset and detailed results are publicly available online, while the platform will remain open for new submissions.

COSMIN methodology for evaluating the content validity of patient-reported outcome measures: a Delphi study

Abstract: Content validity is the most important measurement property of a patient-reported outcome measure (PROM) and the most challenging to assess. Our aims were to: (1) develop standards for evaluating the quality of PROM development; (2) update the original COSMIN standards for assessing the quality of content validity studies of PROMs; (3) develop criteria for what constitutes good content validity of PROMs, and (4) develop a rating system for summarizing the evidence on a PROM’s content validity and grading the quality of the evidence in systematic reviews of PROMs. An online 4-round Delphi study was performed among 159 experts from 21 countries. Panelists rated the degree to which they (dis)agreed to proposed standards, criteria, and rating issues on 5-point rating scales (‘strongly disagree’ to ‘strongly agree’), and provided arguments for their ratings. Discussion focused on sample size requirements, recording and field notes, transcribing cognitive interviews, and data coding. After four rounds, the required 67% consensus was reached on all standards, criteria, and rating issues. After pilot-testing, the steering committee made some final changes. Ten criteria for good content validity were defined regarding item relevance, appropriateness of response options and recall period, comprehensiveness, and comprehensibility of the PROM. The consensus-based COSMIN methodology for content validity is more detailed, standardized, and transparent than earlier published guidelines, including the previous COSMIN standards. This methodology can contribute to the selection and use of high-quality PROMs in research and clinical practice.

Exploring the phenotypic consequences of tissue specific gene expression variation inferred from GWAS summary statistics.

Abstract: Scalable, integrative methods to understand mechanisms that link genetic variants with phenotypes are needed. Here we derive a mathematical expression to compute PrediXcan (a gene mapping approach) results using summary data (S-PrediXcan) and show its accuracy and general robustness to misspecified reference sets. We apply this framework to 44 GTEx tissues and 100+ phenotypes from GWAS and meta-analysis studies, creating a growing public catalog of associations that seeks to capture the effects of gene expression variation on human phenotypes. Replication in an independent cohort is shown. Most of the associations are tissue specific, suggesting context specificity of the trait etiology. Colocalized significant associations in unexpected tissues underscore the need for an agnostic scanning of multiple contexts to improve our ability to detect causal regulatory mechanisms. Monogenic disease genes are enriched among significant associations for related traits, suggesting that smaller alterations of these genes may cause a spectrum of milder phenotypes.

Association analyses of more than 140,000 men identify 63 new prostate cancer susceptibility loci

Abstract: Genome-wide association studies (GWAS) and fine-mapping efforts to date have identified more than 100 prostate cancer (PrCa)-susceptibility loci. We meta-analyzed genotype data from a custom high-density array of 46,939 PrCa cases and 27,910 controls of European ancestry with previously genotyped data of 32,255 PrCa cases and 33,202 controls of European ancestry. Our analysis identified 62 novel loci associated (P C, p.Pro1054Arg) in ATM and rs2066827 (OR = 1.06; P = 2.3 × 10−9; T>G, p.Val109Gly) in CDKN1B. The combination of all loci captured 28.4% of the PrCa familial relative risk, and a polygenic risk score conferred an elevated PrCa risk for men in the ninetieth to ninety-ninth percentiles (relative risk = 2.69; 95% confidence interval (CI): 2.55–2.82) and first percentile (relative risk = 5.71; 95% CI: 5.04–6.48) risk stratum compared with the population average. These findings improve risk prediction, enhance fine-mapping, and provide insight into the underlying biology of PrCa1. A large meta-analysis combining genome-wide and custom high-density genotyping array data identifies 63 new susceptibility loci for prostate cancer, enhancing fine-mapping efforts and providing insights into the underlying biology.

KDEEP: Protein–Ligand Absolute Binding Affinity Prediction via 3D-Convolutional Neural Networks

Abstract: Accurately predicting protein–ligand binding affinities is an important problem in computational chemistry since it can substantially accelerate drug discovery for virtual screening and lead optimization. We propose here a fast machine-learning approach for predicting binding affinities using state-of-the-art 3D-convolutional neural networks and compare this approach to other machine-learning and scoring methods using several diverse data sets. The results for the standard PDBbind (v.2016) core test-set are state-of-the-art with a Pearson’s correlation coefficient of 0.82 and a RMSE of 1.27 in pK units between experimental and predicted affinity, but accuracy is still very sensitive to the specific protein used. KDEEP is made available via PlayMolecule.org for users to test easily their own protein–ligand complexes, with each prediction taking a fraction of a second. We believe that the speed, performance, and ease of use of KDEEP makes it already an attractive scoring function for modern computational ch...

The Beaker phenomenon and the genomic transformation of northwest Europe

Abstract: From around 2750 to 2500 bc, Bell Beaker pottery became widespread across western and central Europe, before it disappeared between 2200 and 1800 bc. The forces that propelled its expansion are a matter of long-standing debate, and there is support for both cultural diffusion and migration having a role in this process. Here we present genome-wide data from 400 Neolithic, Copper Age and Bronze Age Europeans, including 226 individuals associated with Beaker-complex artefacts. We detected limited genetic affinity between Beaker-complex-associated individuals from Iberia and central Europe, and thus exclude migration as an important mechanism of spread between these two regions. However, migration had a key role in the further dissemination of the Beaker complex. We document this phenomenon most clearly in Britain, where the spread of the Beaker complex introduced high levels of steppe-related ancestry and was associated with the replacement of approximately 90% of Britain's gene pool within a few hundred years, continuing the east-to-west expansion that had brought steppe-related ancestry into central and northern Europe over the previous centuries.

Multi-Ethnic Genome-wide Association Study for Atrial Fibrillation

Abstract: Atrial fibrillation (AF) affects more than 33 million individuals worldwide1 and has a complex heritability2. We conducted the largest meta-analysis of genome-wide association studies (GWAS) for AF to date, consisting of more than half a million individuals, including 65,446 with AF. In total, we identified 97 loci significantly associated with AF, including 67 that were novel in a combined-ancestry analysis, and 3 that were novel in a European-specific analysis. We sought to identify AF-associated genes at the GWAS loci by performing RNA-sequencing and expression quantitative trait locus analyses in 101 left atrial samples, the most relevant tissue for AF. We also performed transcriptome-wide analyses that identified 57 AF-associated genes, 42 of which overlap with GWAS loci. The identified loci implicate genes enriched within cardiac developmental, electrophysiological, contractile and structural pathways. These results extend our understanding of the biological pathways underlying AF and may facilitate the development of therapeutics for AF.

The Gut Microbiome Profile in Obesity: A Systematic Review

Abstract: Gut microbiome has been identified in the past decade as an important factor involved in obesity, but the magnitude of its contribution to obesity and its related comorbidities is still uncertain. Among the vast quantity of factors attributed to obesity, environmental, dietary, lifestyle, genetic, and others, the microbiome has aroused curiosity, and the scientific community has published many original articles. Most of the studies related to microbiome and obesity have been reported based on the associations between microbiota and obesity, and the in-depth study of the mechanisms related has been studied mainly in rodents and exceptionally in humans. Due to the quantity and diverse information published, the need of reviews is mandatory to recapitulate the relevant achievements. In this systematic review, we provide an overview of the current evidence on the association between intestinal microbiota and obesity. Additionally, we analyze the effects of an extreme weight loss intervention such as bariatric surgery on gut microbiota. The review is divided into 2 sections: first, the association of obesity and related metabolic disorders with different gut microbiome profiles, including metagenomics studies, and second, changes on gut microbiome after an extreme weight loss intervention such as bariatric surgery.

Towards a complete map of the human long non-coding RNA transcriptome

Abstract: Gene maps, or annotations, enable us to navigate the functional landscape of our genome. They are a resource upon which virtually all studies depend, from single-gene to genome-wide scales and from basic molecular biology to medical genetics. Yet present-day annotations suffer from trade-offs between quality and size, with serious but often unappreciated consequences for downstream studies. This is particularly true for long non-coding RNAs (lncRNAs), which are poorly characterized compared to protein-coding genes. Long-read sequencing technologies promise to improve current annotations, paving the way towards a complete annotation of lncRNAs expressed throughout a human lifetime.

Transcription Factor NRF2 as a Therapeutic Target for Chronic Diseases: A Systems Medicine Approach

Abstract: Systems medicine has a mechanism-based rather than a symptom- or organ-based approach to disease and identifies therapeutic targets in a nonhypothesis-driven manner. In this work, we apply this to transcription factor nuclear factor (erythroid-derived 2)-like 2 (NRF2) by cross-validating its position in a protein-protein interaction network (the NRF2 interactome) functionally linked to cytoprotection in low-grade stress, chronic inflammation, metabolic alterations, and reactive oxygen species formation. Multiscale network analysis of these molecular profiles suggests alterations of NRF2 expression and activity as a common mechanism in a subnetwork of diseases (the NRF2 diseasome). This network joins apparently heterogeneous phenotypes such as autoimmune, respiratory, digestive, cardiovascular, metabolic, and neurodegenerative diseases, along with cancer. Importantly, this approach matches and confirms in silico several applications for NRF2-modulating drugs validated in vivo at different phases of clinical development. Pharmacologically, their profile is as diverse as electrophilic dimethyl fumarate, synthetic triterpenoids like bardoxolone methyl and sulforaphane, protein-protein or DNA-protein interaction inhibitors, and even registered drugs such as metformin and statins, which activate NRF2 and may be repurposed for indications within the NRF2 cluster of disease phenotypes. Thus, NRF2 represents one of the first targets fully embraced by classic and systems medicine approaches to facilitate both drug development and drug repurposing by focusing on a set of disease phenotypes that appear to be mechanistically linked. The resulting NRF2 drugome may therefore rapidly advance several surprising clinical options for this subset of chronic diseases.

A Wavenet for Speech Denoising

Abstract: Most speech processing techniques use magnitude spectrograms as front-end and are therefore by default discarding part of the signal: the phase. In order to overcome this limitation’ we propose an end-to-end learning method for speech denoising based on Wavenet. The proposed model adaptation retains Wavenet's powerful acoustic modeling capabilities, while significantly reducing its time-complexity by eliminating its autoregressive nature. Specifically, the model makes use of non-causal, dilated convolutions and predicts target fields instead of a single target sample. The discriminative adaptation of the model we propose, learns in a supervised fashion via minimizing a regression loss. These modifications make the model highly parallelizable during both training and inference. Both quantitative and qualitative evaluations indicate that the proposed method is preferred over Wiener filtering, a common method based on processing the magnitude spectrogram.

Overcoming catastrophic forgetting with hard attention to the task

Abstract: Catastrophic forgetting occurs when a neural network loses the information learned in a previous task after training on subsequent tasks. This problem remains a hurdle for artificial intelligence systems with sequential learning capabilities. In this paper, we propose a task-based hard attention mechanism that preserves previous tasks' information without affecting the current task's learning. A hard attention mask is learned concurrently to every task, through stochastic gradient descent, and previous masks are exploited to condition such learning. We show that the proposed mechanism is effective for reducing catastrophic forgetting, cutting current rates by 45 to 80%. We also show that it is robust to different hyperparameter choices, and that it offers a number of monitoring capabilities. The approach features the possibility to control both the stability and compactness of the learned knowledge, which we believe makes it also attractive for online learning or network compression applications.

Cancer Genome Interpreter annotates the biological and clinical relevance of tumor alterations

Abstract: While tumor genome sequencing has become widely available in clinical and research settings, the interpretation of tumor somatic variants remains an important bottleneck. Here we present the Cancer Genome Interpreter, a versatile platform that automates the interpretation of newly sequenced cancer genomes, annotating the potential of alterations detected in tumors to act as drivers and their possible effect on treatment response. The results are organized in different levels of evidence according to current knowledge, which we envision can support a broad range of oncology use cases. The resource is publicly available at http://www.cancergenomeinterpreter.org .

Current state of Alzheimer's fluid biomarkers.

Abstract: Alzheimer’s disease (AD) is a progressive neurodegenerative disease with a complex and heterogeneous pathophysiology. The number of people living with AD is predicted to increase; however, there are no disease-modifying therapies currently available and none have been successful in late-stage clinical trials. Fluid biomarkers measured in cerebrospinal fluid (CSF) or blood hold promise for enabling more effective drug development and establishing a more personalized medicine approach for AD diagnosis and treatment. Biomarkers used in drug development programmes should be qualified for a specific context of use (COU). These COUs include, but are not limited to, subject/patient selection, assessment of disease state and/or prognosis, assessment of mechanism of action, dose optimization, drug response monitoring, efficacy maximization, and toxicity/adverse reactions identification and minimization. The core AD CSF biomarkers Aβ42, t-tau, and p-tau are recognized by research guidelines for their diagnostic utility and are being considered for qualification for subject selection in clinical trials. However, there is a need to better understand their potential for other COUs, as well as identify additional fluid biomarkers reflecting other aspects of AD pathophysiology. Several novel fluid biomarkers have been proposed, but their role in AD pathology and their use as AD biomarkers have yet to be validated. In this review, we summarize some of the pathological mechanisms implicated in the sporadic AD and highlight the data for several established and novel fluid biomarkers (including BACE1, TREM2, YKL-40, IP-10, neurogranin, SNAP-25, synaptotagmin, α-synuclein, TDP-43, ferritin, VILIP-1, and NF-L) associated with each mechanism. We discuss the potential COUs for each biomarker.

SUPPA2: fast, accurate, and uncertainty-aware differential splicing analysis across multiple conditions

Abstract: Despite the many approaches to study differential splicing from RNA-seq, many challenges remain unsolved, including computing capacity and sequencing depth requirements. Here we present SUPPA2, a new method that addresses these challenges, and enables streamlined analysis across multiple conditions taking into account biological variability. Using experimental and simulated data, we show that SUPPA2 achieves higher accuracy compared to other methods, especially at low sequencing depth and short read length. We use SUPPA2 to identify novel Transformer2-regulated exons, novel microexons induced during differentiation of bipolar neurons, and novel intron retention events during erythroblast differentiation.

Treatment gap for anxiety disorders is global: Results of the World Mental Health Surveys in 21 countries

Abstract: 1 Background: Anxiety disorders are a major cause of burden of disease. Treatment gaps have been described, but a worldwide evaluation is lacking. We estimated, among individuals with a 12‐month DSM‐IV (where DSM is Diagnostic Statistical Manual) anxiety disorder in 21 countries, the proportion who (i) perceived a need for treatment; (ii) received any treatment; and (iii) received possibly adequate treatment. 2 Methods: Data from 23 community surveys in 21 countries of the World Mental Health (WMH) surveys. DSM‐IV mental disorders were assessed (WHO Composite International Diagnostic Interview, CIDI 3.0). DSM‐IV included posttraumatic stress disorder among anxiety disorders, while it is not considered so in the DSM‐5. We asked if, in the previous 12 months, respondents felt they needed professional treatment and if they obtained professional treatment (specialized/general medical, complementary alternative medical, or nonmedical professional) for “problems with emotions, nerves, mental health, or use of alcohol or drugs.” Possibly adequate treatment was defined as receiving pharmacotherapy (1+ months of medication and 4+ visits to a medical doctor) or psychotherapy, complementary alternative medicine or nonmedical care (8+ visits). 3 Results: Of 51,547 respondents (response = 71.3%), 9.8% had a 12‐month DSM‐IV anxiety disorder, 27.6% of whom received any treatment, and only 9.8% received possibly adequate treatment. Of those with 12‐month anxiety only 41.3% perceived a need for care. Lower treatment levels were found for lower income countries. 4 Conclusions: Low levels of service use and a high proportion of those receiving services not meeting adequacy standards for anxiety disorders exist worldwide. Results suggest the need for improving recognition of anxiety disorders and the quality of treatment.

Transcription factors orchestrate dynamic interplay between genome topology and gene regulation during cell reprogramming

Abstract: Chromosomal architecture is known to influence gene expression, yet its role in controlling cell fate remains poorly understood. Reprogramming of somatic cells into pluripotent stem cells (PSCs) by the transcription factors (TFs) OCT4, SOX2, KLF4 and MYC offers an opportunity to address this question but is severely limited by the low proportion of responding cells. We have recently developed a highly efficient reprogramming protocol that synchronously converts somatic into pluripotent stem cells. Here, we used this system to integrate time-resolved changes in genome topology with gene expression, TF binding and chromatin-state dynamics. The results showed that TFs drive topological genome reorganization at multiple architectural levels, often before changes in gene expression. Removal of locus-specific topological barriers can explain why pluripotency genes are activated sequentially, instead of simultaneously, during reprogramming. Together, our results implicate genome topology as an instructive force for implementing transcriptional programs and cell fate in mammals.

Phenotype, penetrance, and treatment of 133 cytotoxic T-lymphocyte antigen 4–insufficient subjects

Abstract: Background: Cytotoxic T-lymphocyte antigen 4 (CTLA-4) is a negative immune regulator. Heterozygous CTLA4 germline mutations can cause a complex immune dysregulation syndrome in human subjects. Objective: We sought to characterize the penetrance, clinical features, and best treatment options in 133 CTLA4 mutation carriers. Methods: Genetics, clinical features, laboratory values, and outcomes of treatment options were assessed in a worldwide cohort of CTLA4 mutation carriers. Results: We identified 133 subjects from 54 unrelated families carrying 45 different heterozygous CTLA4 mutations, including 28 previously undescribed mutations. Ninety mutation carriers were considered affected, suggesting a clinical penetrance of at least 67%; median age of onset was 11 years, and the mortality rate within affected mutation carriers was 16%(n = 15). Main clinical manifestations included hypogammaglobulinemia (84%), lymphoproliferation (73%), autoimmune cytopenia (62%), and respiratory (68%), gastrointestinal (59%), or neurological features (29%). Eight affectedmutation carriers had lymphoma, and 3 had gastric cancer. An EBV association was found in 6 patients with malignancies. CTLA4 mutations were associated with lymphopenia and decreased T-, B-, and natural killer (NK) cell counts. Successful targeted therapies included application of CTLA-4 fusion proteins, mechanistic target of rapamycin inhibitors, and hematopoietic stem cell transplantation. EBV reactivation occurred in 2 affected mutation carriers after immunosuppression. Conclusions: Affected mutation carriers with CTLA-4 insufficiency can present in any medical specialty. Family members should be counseled because disease manifestation can occur as late as 50 years of age. EBV- and cytomegalovirus-associated complications must be closely monitored. Treatment interventions should be coordinated in clinical trials.

Consistent inverse correlation between DNA methylation of the first intron and gene expression across tissues and species

Abstract: DNA methylation is one of the main epigenetic mechanisms for the regulation of gene expression in eukaryotes. In the standard model, methylation in gene promoters has received the most attention since it is generally associated with transcriptional silencing. Nevertheless, recent studies in human tissues reveal that methylation of the region downstream of the transcription start site is highly informative of gene expression. Also, in some cell types and specific genes it has been found that methylation of the first intron, a gene feature typically rich in enhancers, is linked with gene expression. However, a genome-wide, tissue-independent, systematic comparative analysis of the relationship between DNA methylation in the first intron and gene expression across vertebrates has not been explored yet. The most important findings of this study are: (1) using different tissues from a modern fish, we show a clear genome-wide, tissue-independent quasi-linear inverse relationship between DNA methylation of the first intron and gene expression. (2) This relationship is conserved across vertebrates, since it is also present in the genomes of a model pufferfish, a model frog and different human tissues. Among the gene features, tissues and species interrogated, the first intron’s negative correlation with the gene expression was most consistent. (3) We identified more tissue-specific differentially methylated regions (tDMRs) in the first intron than in any other gene feature. These tDMRs have positive or negative correlation with gene expression, indicative of distinct mechanisms of tissue-specific regulation. (4) Lastly, we identified CpGs in transcription factor binding motifs, enriched in the first intron, the methylation of which tended to increase with the distance from the first exon–first intron boundary, with a concomitant decrease in gene expression. Our integrative analysis clearly reveals the important and conserved role of the methylation level of the first intron and its inverse association with gene expression regardless of tissue and species. These findings not only contribute to our basic understanding of the epigenetic regulation of gene expression but also identify the first intron as an informative gene feature regarding the relationship between DNA methylation and gene expression where future studies should be focused.

Overcoming catastrophic forgetting with hard attention to the task

Abstract: Comunicacio presentada a: 35th International Conference on Machine Learning, celebrat a Stockholmsmassan, Suecia, del 10 al 15 de juliol del 2018.

Comparison of Perioperative Outcomes Between Laparoscopic and Open Approach for Pancreatoduodenectomy: The PADULAP Randomized Controlled Trial.

Abstract: Objective:To compare perioperative outcomes of pancreatoduodenectomy (PD) performed through the laparoscopic route or by open surgery.Summary Background Data:Laparoscopic PD is being progressively performed in selected patients.Methods:An open-label single-center RCT was conducted between February 2