Search

Cancer driver genes, i.e., oncogenes and tumor suppressor genes, are involved in the acquisition of important functions in tumors, providing a selective growth advantage, allowing uncontrolled proliferation and avoiding apoptosis. It is therefore important to identify these driver genes, both for the fundamental understanding of cancer and to help finding new therapeutic targets or biomarkers. Although the most frequently mutated driver genes have been identified, it is believed that many more remain to be discovered, particularly for driver genes specific to some cancer types. In this paper, we propose a new computational method called LOTUS to predict new driver genes. LOTUS is a machine-learning based approach which allows to integrate various types of data in a versatile manner, including information about gene mutations and protein-protein interactions. In addition, LOTUS can predict cancer driver genes in a pan-cancer setting as well as for specific cancer types, using a multitask learning strategy to share information across cancer types. We empirically show that LOTUS outperforms five other state-of-the-art driver gene prediction methods, both in terms of intrinsic consistency and prediction accuracy, and provide predictions of new cancer genes across many cancer types. [ABSTRACT FROM AUTHOR]

Robustly predicting outcome for cancer patients from gene expression is an important challenge on the road to better personalized treatment. Network-based outcome predictors (NOPs), which considers the cellular wiring diagram in the classification, hold much promise to improve performance, stability and interpretability of identified marker genes. Problematically, reports on the efficacy of NOPs are conflicting and for instance suggest that utilizing random networks performs on par to networks that describe biologically relevant interactions. In this paper we turn the prediction problem around: instead of using a given biological network in the NOP, we aim to identify the network of genes that truly improves outcome prediction. To this end, we propose SyNet, a gene network constructed ab initio from synergistic gene pairs derived from survival-labelled gene expression data. To obtain SyNet, we evaluate synergy for all 69 million pairwise combinations of genes resulting in a network that is specific to the dataset and phenotype under study and can be used to in a NOP model. We evaluated SyNet and 11 other networks on a compendium dataset of >4000 survival-labelled breast cancer samples. For this purpose, we used cross-study validation which more closely emulates real world application of these outcome predictors. We find that SyNet is the only network that truly improves performance, stability and interpretability in several existing NOPs. We show that SyNet overlaps significantly with existing gene networks, and can be confidently predicted (~85% AUC) from graph-topological descriptions of these networks, in particular the breast tissue-specific network. Due to its data-driven nature, SyNet is not biased to well-studied genes and thus facilitates post-hoc interpretation. We find that SyNet is highly enriched for known breast cancer genes and genes related to e.g. histological grade and tamoxifen resistance, suggestive of a role in determining breast cancer outcome. [ABSTRACT FROM AUTHOR]

LncRNA-protein interactions play important roles in post-transcriptional gene regulation, poly-adenylation, splicing and translation. Identification of lncRNA-protein interactions helps to understand lncRNA-related activities. Existing computational methods utilize multiple lncRNA features or multiple protein features to predict lncRNA-protein interactions, but features are not available for all lncRNAs or proteins; most of existing methods are not capable of predicting interacting proteins (or lncRNAs) for new lncRNAs (or proteins), which don’t have known interactions. In this paper, we propose the sequence-based feature projection ensemble learning method, “SFPEL-LPI”, to predict lncRNA-protein interactions. First, SFPEL-LPI extracts lncRNA sequence-based features and protein sequence-based features. Second, SFPEL-LPI calculates multiple lncRNA-lncRNA similarities and protein-protein similarities by using lncRNA sequences, protein sequences and known lncRNA-protein interactions. Then, SFPEL-LPI combines multiple similarities and multiple features with a feature projection ensemble learning frame. In computational experiments, SFPEL-LPI accurately predicts lncRNA-protein associations and outperforms other state-of-the-art methods. More importantly, SFPEL-LPI can be applied to new lncRNAs (or proteins). The case studies demonstrate that our method can find out novel lncRNA-protein interactions, which are confirmed by literature. Finally, we construct a user-friendly web server, available at . [ABSTRACT FROM AUTHOR]

Classification models built on biological expression data are increasingly used to predict distinct disease subtypes. Selected features that separate sample groups can be the candidates of biomarkers, helping us to discover biological functions/pathways. However, three challenges are associated with building a robust classification and feature selection model: 1) the number of significant biomarkers is much smaller than that of measured features for which the search will be exhaustive; 2) current biological expression data are big in both sample size and feature size which will worsen the scalability of any search algorithms; and 3) expression profiles of certain features are typically highly correlated which may prevent to distinguish the predominant features. Unfortunately, most of the existing algorithms are partially addressing part of these challenges but not as a whole. In this paper, we propose a unified framework to address the above challenges. The classification and feature selection problem is first formulated as a nonconvex optimisation problem. Then the problem is relaxed and solved iteratively by a sequence of convex optimisation procedures which can be distributed computed and therefore allows the efficient implementation on advanced infrastructures. To illustrate the competence of our method over others, we first analyse a randomly generated simulation dataset under various conditions. We then analyse a real gene expression dataset on embryonal tumour. Further downstream analysis, such as functional annotation and pathway analysis, are performed on the selected features which elucidate several biological findings. [ABSTRACT FROM AUTHOR]

Human endogenous retroviruses (HERVs) encode active retroviral proteins, which may be involved in the progression of cancer and other diseases. Matrix protein (MA), in group-specific antigen genes (gag) of retroviruses, is associated with the virus envelope glycoproteins in most mammalian retroviruses and may be involved in virus particle assembly, transport and budding. However, the amount of annotated MAs in ERVs is still at a low level so far. No computational method to predict the exact start and end coordinates of MAs in gags has been proposed yet. In this paper, a computational method to identify MAs in ERVs is proposed. A divide and conquer technique was designed and applied to the conventional prediction model to acquire better results when dealing with gene sequences with various lengths. Initiation sites and termination sites were predicted separately and then combined according to their intervals. Three different algorithms were applied and compared: weighted support vector machine (WSVM), weighted extreme learning machine (WELM) and random forest (RF). G − mean (geometric mean of sensitivity and specificity) values of initiation sites and termination sites under 5-fold cross validation generated by random forest models are 0.9869 and 0.9755 respectively, highest among the algorithms applied. Our prediction models combine RF & WSVM algorithms to achieve the best prediction results. 98.4% of all the collected ERV sequences with complete MAs (125 in total) could be predicted exactly correct by the models. 94,671 HERV sequences from 118 families were scanned by the model, 104 new putative MAs were predicted in human chromosomes. Distributions of the putative MAs and optimizations of model parameters were also analyzed. The usage of our predicting method was also expanded to other retroviruses and satisfying results were acquired. [ABSTRACT FROM AUTHOR]

Genes that are indispensable for survival are essential genes. Many features have been proposed for computational prediction of essential genes. In this paper, the least absolute shrinkage and selection operator method was used to screen key sequence-based features related to gene essentiality. To assess the effects, the selected features were used to predict the essential genes from 31 bacterial species based on a support vector machine classifier. For all 31 bacterial objects (21 Gram-negative objects and ten Gram-positive objects), the features in the three datasets were reduced from 57, 59, and 58, to 40, 37, and 38, respectively, without loss of prediction accuracy. Results showed that some features were redundant for gene essentiality, so could be eliminated from future analyses. The selected features contained more complex (or key) biological information for gene essentiality, and could be of use in related research projects, such as gene prediction, synthetic biology, and drug design. [ABSTRACT FROM AUTHOR]

Among the scientific challenges posed by complex diseases with a strong genetic component, two stand out. One is unveiling the role of rare and common genetic variants; the other is the design of classification models to improve clinical diagnosis and predictive models for prognosis and personalized therapies. In this paper, we present a data fusion framework merging gene, domain, pathway and protein-protein interaction data related to a next generation sequencing epilepsy gene panel. Our method allows integrating association information from multiple genomic sources and aims at highlighting the set of common and rare variants that are capable to trigger the occurrence of a complex disease. When compared to other approaches, our method shows better performances in classifying patients affected by epilepsy. [ABSTRACT FROM AUTHOR]

Genetic risk prediction is an important goal in human genetics research and precision medicine. Accurate prediction models will have great impacts on both disease prevention and early treatment strategies. Despite the identification of thousands of disease-associated genetic variants through genome wide association studies (GWAS), genetic risk prediction accuracy remains moderate for most diseases, which is largely due to the challenges in both identifying all the functionally relevant variants and accurately estimating their effect sizes in the presence of linkage disequilibrium. In this paper, we introduce AnnoPred, a principled framework that leverages diverse types of genomic and epigenomic functional annotations in genetic risk prediction for complex diseases. AnnoPred is trained using GWAS summary statistics in a Bayesian framework in which we explicitly model various functional annotations and allow for linkage disequilibrium estimated from reference genotype data. Compared with state-of-the-art risk prediction methods, AnnoPred achieves consistently improved prediction accuracy in both extensive simulations and real data. [ABSTRACT FROM AUTHOR]

With the increasing availability of high-throughput data, various computational methods have recently been developed for understanding the cell through protein-protein interaction (PPI) networks at a systems level. However, due to the incompleteness of the original PPI networks those efforts have been significantly hindered. In this paper, we propose a two stage method to predict underlying links between two originally unlinked protein pairs. First, we measure gene expression and gene functional similarly between unlinked protein pairs on Saccharomyces cerevisiae benchmark network and obtain new constructed networks. Then, we select the significant part of the new predicted links by analyzing the difference between essential proteins that have been identified based on the new constructed networks and the original network. Furthermore, we validate the performance of the new method by using the reliable and comprehensive PPI dataset obtained from the STRING database and compare the new proposed method with four other random walk-based methods. Comparing the results indicates that the new proposed strategy performs well in predicting underlying links. This study provides a general paradigm for predicting new interactions between protein pairs and offers new insights into identifying essential proteins. [ABSTRACT FROM AUTHOR]

Protein-protein interaction network data provides valuable information that infers direct links between genes and their biological roles. This information brings a fundamental hypothesis for protein function prediction that interacting proteins tend to have similar functions. With the help of recently-developed network embedding feature generation methods and deep maxout neural networks, it is possible to extract functional representations that encode direct links between protein-protein interactions information and protein function. Our novel method, STRING2GO, successfully adopts deep maxout neural networks to learn functional representations simultaneously encoding both protein-protein interactions and functional predictive information. The experimental results show that STRING2GO outperforms other protein-protein interaction network-based prediction methods and one benchmark method adopted in a recent large scale protein function prediction competition. [ABSTRACT FROM AUTHOR]

Network medicine approaches have been largely successful at increasing our knowledge of molecularly characterized diseases. Given a set of disease genes associated with a disease, neighbourhood-based methods and random walkers exploit the interactome allowing the prediction of further genes for that disease. In general, however, diseases with no known molecular basis constitute a challenge. Here we present a novel network approach to prioritize gene-disease associations that is able to also predict genes for diseases with no known molecular basis. Our method, which we have called Cardigan (ChARting DIsease Gene AssociatioNs), uses semi-supervised learning and exploits a measure of similarity between disease phenotypes. We evaluated its performance at predicting genes for both molecularly characterized and uncharacterized diseases in OMIM, using both weighted and binary interactomes, and compared it with state-of-the-art methods. Our tests, which use datasets collected at different points in time to replicate the dynamics of the disease gene discovery process, prove that Cardigan is able to accurately predict disease genes for molecularly uncharacterized diseases. Additionally, standard leave-one-out cross validation tests show how our approach outperforms state-of-the-art methods at predicting genes for molecularly characterized diseases by 14%-65%. Cardigan can also be used for disease module prediction, where it outperforms state-of-the-art methods by 87%-299%. [ABSTRACT FROM AUTHOR]

The high concentration of arsenic (As) in rice grains, in a large proportion of the rice growing areas, is a critical issue. This study explores the feasibility of conventional (QTL-based) marker-assisted selection and genomic selection to improve the ability of rice to prevent As uptake and accumulation in the edible grains. A japonica diversity panel (RP) of 228 accessions phenotyped for As concentration in the flag leaf (FL-As) and in the dehulled grain (CG-As), and genotyped at 22,370 SNP loci, was used to map QTLs by association analysis (GWAS) and to train genomic prediction models. Similar phenotypic and genotypic data from 95 advanced breeding lines (VP) with japonica genetic backgrounds, was used to validate related QTLs mapped in the RP through GWAS and to evaluate the predictive ability of across populations (RP-VP) genomic estimate of breeding value (GEBV) for As exclusion. Several QTLs for FL-As and CG-As with a low-medium individual effect were detected in the RP, of which some colocalized with known QTLs and candidate genes. However, less than 10% of those QTLs could be validated in the VP without loosening colocalization parameters. Conversely, the average predictive ability of across populations GEBV was rather high, 0.43 for FL-As and 0.48 for CG-As, ensuring genetic gains per time unit close to phenotypic selection. The implications of the limited robustness of the GWAS results and the rather high predictive ability of genomic prediction are discussed for breeding rice for significantly low arsenic uptake and accumulation in the edible grains. [ABSTRACT FROM AUTHOR]

Developing high yielding rice varieties that are tolerant to drought stress is crucial for the sustainable livelihood of rice farmers in rainfed rice cropping ecosystems. Genomic selection (GS) promises to be an effective breeding option for these complex traits. We evaluated the effectiveness of two rather new options in the implementation of GS: trait and environment-specific marker selection and the use of multi-environment prediction models. A reference population of 280 rainfed lowland accessions endowed with 215k SNP markers data was phenotyped under a favorable and two managed drought environments. Trait-specific SNP subsets (28k) were selected for each trait under each environment, using results of GWAS performed with the complete genotype dataset. Performances of single-environment and multi-environment genomic prediction models were compared using kernel regression based methods (GBLUP and RKHS) under two cross validation scenarios: availability (CV2) or not (CV1) of phenotypic data for the validation set, in one of the environments. Trait-specific marker selection strategy achieved predictive ability (PA) of genomic prediction up to 22% higher than markers selected on the bases of neutral linkage disequilibrium (LD). Tolerance to drought stress was up to 32% better predicted by multi-environment models (especially RKHS based models) under CV2 strategy. Under the less favorable CV1 strategy, the multi-environment models achieved similar PA than the single-environment predictions. We also showed that reasonable PA could be obtained with as few as 3,000 SNP markers, even in a population of low LD extent, provided marker selection is based on pairwise LD. The implications of these findings for breeding for drought tolerance are discussed. The most resource sparing option would be accurate phenotyping of the reference population in a favorable environment and under a managed drought, while the candidate population would be phenotyped only under one of those environments. [ABSTRACT FROM AUTHOR]

Integration of genome-wide association studies (GWAS) and expression quantitative trait loci (eQTL) studies is needed to improve our understanding of the biological mechanisms underlying GWAS hits, and our ability to identify therapeutic targets. Gene-level association methods such as PrediXcan can prioritize candidate targets. However, limited eQTL sample sizes and absence of relevant developmental and disease context restrict our ability to detect associations. Here we propose an efficient statistical method (MultiXcan) that leverages the substantial sharing of eQTLs across tissues and contexts to improve our ability to identify potential target genes. MultiXcan integrates evidence across multiple panels using multivariate regression, which naturally takes into account the correlation structure. We apply our method to simulated and real traits from the UK Biobank and show that, in realistic settings, we can detect a larger set of significantly associated genes than using each panel separately. To improve applicability, we developed a summary result-based extension called S-MultiXcan, which we show yields highly concordant results with the individual level version when LD is well matched. Our multivariate model-based approach allowed us to use the individual level results as a gold standard to calibrate and develop a robust implementation of the summary-based extension. Results from our analysis as well as software and necessary resources to apply our method are publicly available. [ABSTRACT FROM AUTHOR]

We have developed an easy-to-use and memory-efficient method called PhenotypeSeeker that (a) identifies phenotype-specific k-mers, (b) generates a k-mer-based statistical model for predicting a given phenotype and (c) predicts the phenotype from the sequencing data of a given bacterial isolate. The method was validated on 167 Klebsiella pneumoniae isolates (virulence), 200 Pseudomonas aeruginosa isolates (ciprofloxacin resistance) and 459 Clostridium difficile isolates (azithromycin resistance). The phenotype prediction models trained from these datasets obtained the F1-measure of 0.88 on the K. pneumoniae test set, 0.88 on the P. aeruginosa test set and 0.97 on the C. difficile test set. The F1-measures were the same for assembled sequences and raw sequencing data; however, building the model from assembled genomes is significantly faster. On these datasets, the model building on a mid-range Linux server takes approximately 3 to 5 hours per phenotype if assembled genomes are used and 10 hours per phenotype if raw sequencing data are used. The phenotype prediction from assembled genomes takes less than one second per isolate. Thus, PhenotypeSeeker should be well-suited for predicting phenotypes from large sequencing datasets. PhenotypeSeeker is implemented in Python programming language, is open-source software and is available at GitHub (). [ABSTRACT FROM AUTHOR]

Covariate selection is a fundamental step when building sparse prediction models in order to avoid overfitting and to gain a better interpretation of the classifier without losing its predictive accuracy. In practice the LASSO regression of Tibshirani, which penalizes the likelihood of the model by the L1 norm of the regression coefficients, has become the gold-standard to reach these objectives. Recently Lee and Oh developed a novel random-effect covariate selection method called the modified unbounded penalty (MUB) regression, whose penalization function can equal minus infinity at 0 in order to produce very sparse models. We sought to compare the predictive accuracy and the number of covariates selected by these two methods in several high-dimensional datasets, consisting in genes expressions measured to predict response to chemotherapy in breast cancer patients. These comparisons were performed by building the Receiver Operating Characteristics (ROC) curves of the classifiers obtained with the selected genes and by comparing their area under the ROC curve (AUC) corrected for optimism using several variants of bootstrap internal validation and cross-validation. We found consistently in all datasets that the MUB penalization selected a remarkably smaller number of covariates than the LASSO while offering a similar—and encouraging—predictive accuracy. The models selected by the MUB were actually nested in the ones obtained with the LASSO. Similar findings were observed when comparing these results to those obtained in their first publication by other authors or when using the area under the Precision-Recall curve (AUCPR) as another measure of predictive performance. In conclusion, the MUB penalization seems therefore to be one of the best options when sparsity is required in high-dimension. Further investigation in other datasets is however required to validate these findings. [ABSTRACT FROM AUTHOR]

Machine learning methods for protein function prediction are urgently needed, especially now that a substantial fraction of known sequences remains unannotated despite the extensive use of functional assignments based on sequence similarity. One major bottleneck supervised learning faces in protein function prediction is the structured, multi-label nature of the problem, because biological roles are represented by lists of terms from hierarchically organised controlled vocabularies such as the Gene Ontology. In this work, we build on recent developments in the area of deep learning and investigate the usefulness of multi-task deep neural networks (MTDNN), which consist of upstream shared layers upon which are stacked in parallel as many independent modules (additional hidden layers with their own output units) as the number of output GO terms (the tasks). MTDNN learns individual tasks partially using shared representations and partially from task-specific characteristics. When no close homologues with experimentally validated functions can be identified, MTDNN gives more accurate predictions than baseline methods based on annotation frequencies in public databases or homology transfers. More importantly, the results show that MTDNN binary classification accuracy is higher than alternative machine learning-based methods that do not exploit commonalities and differences among prediction tasks. Interestingly, compared with a single-task predictor, the performance improvement is not linearly correlated with the number of tasks in MTDNN, but medium size models provide more improvement in our case. One of advantages of MTDNN is that given a set of features, there is no requirement for MTDNN to have a bootstrap feature selection procedure as what traditional machine learning algorithms do. Overall, the results indicate that the proposed MTDNN algorithm improves the performance of protein function prediction. On the other hand, there is still large room for deep learning techniques to further enhance prediction ability. [ABSTRACT FROM AUTHOR]

Any single nucleotide variant detection study could benefit from a fast and cheap method of measuring the quality of variant call list. It is advantageous to be able to see how the call list quality is affected by different variant filtering thresholds and other adjustments to the study parameters. Here we look into a possibility of estimating the proportion of true positives in a single nucleotide variant call list for human data. Using whole-exome and whole-genome gold standard data sets for training, we focus on building a generic model that only relies on information available from any variant caller. We assess and compare the performance of different candidate models based on their practical accuracy. We find that the generic model delivers decent accuracy most of the time. Further, we conclude that its performance could be improved substantially by leveraging the variant quality metrics that are specific to each variant calling tool. [ABSTRACT FROM AUTHOR]

Hepatitis B virus (HBV) infection is a common problem in the world, especially in China. More than 60–80% of hepatocellular carcinoma (HCC) cases can be attributed to HBV infection in high HBV prevalent regions. Although traditional Sanger sequencing has been extensively used to investigate HBV sequences, NGS is becoming more commonly used. Further, it is unknown whether word pattern frequencies of HBV reads by Next Generation Sequencing (NGS) can be used to investigate HBV genotypes and predict HCC status. In this study, we used NGS to sequence the pre-S region of the HBV sequence of 94 HCC patients and 45 chronic HBV (CHB) infected individuals. Word pattern frequencies among the sequence data of all individuals were calculated and compared using the Manhattan distance. The individuals were grouped using principal coordinate analysis (PCoA) and hierarchical clustering. Word pattern frequencies were also used to build prediction models for HCC status using both K-nearest neighbors (KNN) and support vector machine (SVM). We showed the extremely high power of analyzing HBV sequences using word patterns. Our key findings include that the first principal coordinate of the PCoA analysis was highly associated with the fraction of genotype B (or C) sequences and the second principal coordinate was significantly associated with the probability of having HCC. Hierarchical clustering first groups the individuals according to their major genotypes followed by their HCC status. Using cross-validation, high area under the receiver operational characteristic curve (AUC) of around 0.88 for KNN and 0.92 for SVM were obtained. In the independent data set of 46 HCC patients and 31 CHB individuals, a good AUC score of 0.77 was obtained using SVM. It was further shown that 3000 reads for each individual can yield stable prediction results for SVM. Thus, another key finding is that word patterns can be used to predict HCC status with high accuracy. Therefore, our study shows clearly that word pattern frequencies of HBV sequences contain much information about the composition of different HBV genotypes and the HCC status of an individual. [ABSTRACT FROM AUTHOR]

Building prediction models based on complex omics datasets such as transcriptomics, proteomics, metabolomics remains a challenge in bioinformatics and biostatistics. Regularized regression techniques are typically used to deal with the high dimensionality of these datasets. However, due to the presence of correlation in the datasets, it is difficult to select the best model and application of these methods yields unstable results. We propose a novel strategy for model selection where the obtained models also perform well in terms of overall predictability. Several three step approaches are considered, where the steps are 1) network construction, 2) clustering to empirically derive modules or pathways, and 3) building a prediction model incorporating the information on the modules. For the first step, we use weighted correlation networks and Gaussian graphical modelling. Identification of groups of features is performed by hierarchical clustering. The grouping information is included in the prediction model by using group-based variable selection or group-specific penalization. We compare the performance of our new approaches with standard regularized regression via simulations. Based on these results we provide recommendations for selecting a strategy for building a prediction model given the specific goal of the analysis and the sizes of the datasets. Finally we illustrate the advantages of our approach by application of the methodology to two problems, namely prediction of body mass index in the DIetary, Lifestyle, and Genetic determinants of Obesity and Metabolic syndrome study (DILGOM) and prediction of response of each breast cancer cell line to treatment with specific drugs using a breast cancer cell lines pharmacogenomics dataset. [ABSTRACT FROM AUTHOR]

The adaptation of the CRISPR-Cas9 system as a genome editing technique has generated much excitement in recent years owing to its ability to manipulate targeted genes and genomic regions that are complementary to a programmed single guide RNA (sgRNA). However, the efficacy of a specific sgRNA is not uniquely defined by exact sequence homology to the target site, thus unintended off-targets might additionally be cleaved. Current methods for sgRNA design are mainly concerned with predicting off-targets for a given sgRNA using basic sequence features and employ elementary rules for ranking possible sgRNAs. Here, we introduce CRISTA (CRISPR Target Assessment), a novel algorithm within the machine learning framework that determines the propensity of a genomic site to be cleaved by a given sgRNA. We show that the predictions made with CRISTA are more accurate than other available methodologies. We further demonstrate that the occurrence of bulges is not a rare phenomenon and should be accounted for in the prediction process. Beyond predicting cleavage efficiencies, the learning process provides inferences regarding patterns that underlie the mechanism of action of the CRISPR-Cas9 system. We discover that attributes that describe the spatial structure and rigidity of the entire genomic site as well as those surrounding the PAM region are a major component of the prediction capabilities. [ABSTRACT FROM AUTHOR]

Recent heritability analyses have indicated that genome-wide association studies (GWAS) have the potential to improve genetic risk prediction for complex diseases based on polygenic risk score (PRS), a simple modelling technique that can be implemented using summary-level data from the discovery samples. We herein propose modifications to improve the performance of PRS. We introduce threshold-dependent winner’s-curse adjustments for marginal association coefficients that are used to weight the single-nucleotide polymorphisms (SNPs) in PRS. Further, as a way to incorporate external functional/annotation knowledge that could identify subsets of SNPs highly enriched for associations, we propose variable thresholds for SNPs selection. We applied our methods to GWAS summary-level data of 14 complex diseases. Across all diseases, a simple winner’s curse correction uniformly led to enhancement of performance of the models, whereas incorporation of functional SNPs was beneficial only for selected diseases. Compared to the standard PRS algorithm, the proposed methods in combination led to notable gain in efficiency (25–50% increase in the prediction R2) for 5 of 14 diseases. As an example, for GWAS of type 2 diabetes, winner’s curse correction improved prediction R2 from 2.29% based on the standard PRS to 3.10% (P = 0.0017) and incorporating functional annotation data further improved R2 to 3.53% (P = 2×10−5). Our simulation studies illustrate why differential treatment of certain categories of functional SNPs, even when shown to be highly enriched for GWAS-heritability, does not lead to proportionate improvement in genetic risk-prediction because of non-uniform linkage disequilibrium structure. [ABSTRACT FROM AUTHOR]

Transcript levels do not faithfully predict protein levels, due to post-transcriptional regulation of gene expression mediated by RNA binding proteins (RBPs) and non-coding RNAs. We developed a multivariate linear regression model integrating RBP levels and predicted RBP-mRNA regulatory interactions from matched transcript and protein datasets. RBPs significantly improved the accuracy in predicting protein abundance of a portion of the total modeled mRNAs in three panels of tissues and cells and for different methods employed in the detection of mRNA and protein. The presence of upstream translation initiation sites (uTISs) at the mRNA 5’ untranslated regions was strongly associated with improvement in predictive accuracy. On the basis of these observations, we propose that the recently discovered widespread uTISs in the human genome can be a previously unappreciated substrate of translational control mediated by RBPs. [ABSTRACT FROM AUTHOR]

Understanding the genetic architecture of gene expression traits is key to elucidating the underlying mechanisms of complex traits. Here, for the first time, we perform a systematic survey of the heritability and the distribution of effect sizes across all representative tissues in the human body. We find that local h2 can be relatively well characterized with 59% of expressed genes showing significant h2 (FDR < 0.1) in the DGN whole blood cohort. However, current sample sizes (n ≤ 922) do not allow us to compute distal h2. Bayesian Sparse Linear Mixed Model (BSLMM) analysis provides strong evidence that the genetic contribution to local expression traits is dominated by a handful of genetic variants rather than by the collective contribution of a large number of variants each of modest size. In other words, the local architecture of gene expression traits is sparse rather than polygenic across all 40 tissues (from DGN and GTEx) examined. This result is confirmed by the sparsity of optimal performing gene expression predictors via elastic net modeling. To further explore the tissue context specificity, we decompose the expression traits into cross-tissue and tissue-specific components using a novel Orthogonal Tissue Decomposition (OTD) approach. Through a series of simulations we show that the cross-tissue and tissue-specific components are identifiable via OTD. Heritability and sparsity estimates of these derived expression phenotypes show similar characteristics to the original traits. Consistent properties relative to prior GTEx multi-tissue analysis results suggest that these traits reflect the expected biology. Finally, we apply this knowledge to develop prediction models of gene expression traits for all tissues. The prediction models, heritability, and prediction performance R2 for original and decomposed expression phenotypes are made publicly available (). [ABSTRACT FROM AUTHOR]

Target-based screening is one of the major approaches in drug discovery. Besides the intended target, unexpected drug off-target interactions often occur, and many of them have not been recognized and characterized. The off-target interactions can be responsible for either therapeutic or side effects. Thus, identifying the genome-wide off-targets of lead compounds or existing drugs will be critical for designing effective and safe drugs, and providing new opportunities for drug repurposing. Although many computational methods have been developed to predict drug-target interactions, they are either less accurate than the one that we are proposing here or computationally too intensive, thereby limiting their capability for large-scale off-target identification. In addition, the performances of most machine learning based algorithms have been mainly evaluated to predict off-target interactions in the same gene family for hundreds of chemicals. It is not clear how these algorithms perform in terms of detecting off-targets across gene families on a proteome scale. Here, we are presenting a fast and accurate off-target prediction method, REMAP, which is based on a dual regularized one-class collaborative filtering algorithm, to explore continuous chemical space, protein space, and their interactome on a large scale. When tested in a reliable, extensive, and cross-gene family benchmark, REMAP outperforms the state-of-the-art methods. Furthermore, REMAP is highly scalable. It can screen a dataset of 200 thousands chemicals against 20 thousands proteins within 2 hours. Using the reconstructed genome-wide target profile as the fingerprint of a chemical compound, we predicted that seven FDA-approved drugs can be repurposed as novel anti-cancer therapies. The anti-cancer activity of six of them is supported by experimental evidences. Thus, REMAP is a valuable addition to the existing in silico toolbox for drug target identification, drug repurposing, phenotypic screening, and side effect prediction. The software and benchmark are available at . [ABSTRACT FROM AUTHOR]

3D-domain swapping is one of the mechanisms of protein oligomerization and the proteins exhibiting this phenomenon have many biological functions. These proteins, which undergo domain swapping, have acquired much attention owing to their involvement in human diseases, such as conformational diseases, amyloidosis, serpinopathies, proteionopathies etc. Early realisation of proteins in the whole human genome that retain tendency to domain swap will enable many aspects of disease control management. Predictive models were developed by using machine learning approaches with an average accuracy of 78% (85.6% of sensitivity, 87.5% of specificity and an MCC value of 0.72) to predict putative domain swapping in protein sequences. These models were applied to many complete genomes with special emphasis on the human genome. Nearly 44% of the protein sequences in the human genome were predicted positive for domain swapping. Enrichment analysis was performed on the positively predicted sequences from human genome for their domain distribution, disease association and functional importance based on Gene Ontology (GO). Enrichment analysis was also performed to infer a better understanding of the functional importance of these sequences. Finally, we developed hinge region prediction, in the given putative domain swapped sequence, by using important physicochemical properties of amino acids. [ABSTRACT FROM AUTHOR]

Genomic selection is focused on prediction of breeding values of selection candidates by means of high density of markers. It relies on the assumption that all quantitative trait loci (QTLs) tend to be in strong linkage disequilibrium (LD) with at least one marker. In this context, we present theoretical results regarding the accuracy of genomic selection, i.e., the correlation between predicted and true breeding values. Typically, for individuals (so-called test individuals), breeding values are predicted by means of markers, using marker effects estimated by fitting a ridge regression model to a set of training individuals. We present a theoretical expression for the accuracy; this expression is suitable for any configurations of LD between QTLs and markers. We also introduce a new accuracy proxy that is free of the QTL parameters and easily computable; it outperforms the proxies suggested in the literature, in particular, those based on an estimated effective number of independent loci (Me). The theoretical formula, the new proxy, and existing proxies were compared for simulated data, and the results point to the validity of our approach. The calculations were also illustrated on a new perennial ryegrass set (367 individuals) genotyped for 24,957 single nucleotide polymorphisms (SNPs). In this case, most of the proxies studied yielded similar results because of the lack of markers for coverage of the entire genome (2.7 Gb). [ABSTRACT FROM AUTHOR]

Systems biology offers promising approaches for identifying response-specific signatures to vaccination and assessing their predictive value. Here, we designed a modelling strategy aiming to predict the quality of late T-cell responses after vaccination from early transcriptome analysis of dendritic cells. Using standardized staining with tetramer, we first quantified antigen-specific T-cell expansion 5 to 10 days after vaccination with one of a set of 41 different vaccine vectors all expressing the same antigen. Hierarchical clustering of the responses defined sets of high and low T cell response inducers. We then compared these responses with the transcriptome of splenic dendritic cells obtained 6 hours after vaccination with the same vectors and produced a random forest model capable of predicting the quality of the later antigen-specific T-cell expansion. The model also successfully predicted vector classification as low or strong T-cell response inducers of a novel set of vaccine vectors, based on the early transcriptome results obtained from spleen dendritic cells, whole spleen and even peripheral blood mononuclear cells. Finally, our model developed with mouse datasets also accurately predicted vaccine efficacy from literature-mined human datasets. [ABSTRACT FROM AUTHOR]