A New Machine Learning-Based Framework for Mapping Uncertainty Analysis in RNA-Seq Read Alignment and Gene Expression Estimation

DifferentialRegulation: a Bayesian hierarchical approach to identify differentially regulated genes

Although transcriptomics data is typically used to analyze mature spliced mRNA, recent attention has focused on jointly investigating spliced and unspliced (or precursor-) mRNA, which can be used to study gene regulation and changes in gene expression production. Nonetheless, most methods for spliced/unspliced inference (such as RNA velocity tools) focus on individual samples, and rarely allow comparisons between groups of samples (e.g. healthy vs. diseased). Furthermore, this kind of inference is challenging, because spliced and unspliced mRNA abundance is characterized by a high degree of quantification uncertainty, due to the prevalence of multi-mapping reads, ie reads compatible with multiple transcripts (or genes), and/or with both their spliced and unspliced versions. Here, we present DifferentialRegulation, a Bayesian hierarchical method to discover changes between experimental conditions with respect to the relative abundance of unspliced mRNA (over the total mRNA). We model the quantification uncertainty via a latent variable approach, where reads are allocated to their gene/transcript of origin, and to the respective splice version. We designed several benchmarks where our approach shows good performance, in terms of sensitivity and error control, vs. state-of-the-art competitors. Importantly, our tool is flexible, and works with both bulk and single-cell RNA-sequencing data. DifferentialRegulation is distributed as a Bioconductor R package.

Modeling of astaxanthin biosynthesis via machine learning, mathematical and metabolic network modeling

Natural astaxanthin is synthesized by diverse organisms including: bacteria, fungi, microalgae, and plants involving complex cellular processes, which depend on numerous interrelated parameters. Nonetheless, existing knowledge regarding astaxanthin biosynthesis and the conditions influencing astaxanthin accumulation is fairly limited. Thus, manipulation of the growth conditions to achieve desired biomass and astaxanthin yields can be a complicated process requiring cost-intensive and time-consuming experiment-based research. As a potential solution, modeling and simulation of biological systems have recently emerged, allowing researchers to predict/estimate astaxanthin production dynamics in selected organisms. Moreover, mathematical modeling techniques would enable further optimization of astaxanthin synthesis in a shorter period of time, ultimately contributing to a notable reduction in production costs. Thus, the present review comprehensively discusses existing mathematical modeling techniques which simulate the bioaccumulation of astaxanthin in diverse organisms. Associated challenges, solutions, and future perspectives are critically analyzed and presented.

De novo assembly of transcriptomes and differential gene expression analysis using short-read data from emerging model organisms - a brief guide

Many questions in biology benefit greatly from the use of a variety of model systems. High-throughput sequencing methods have been a triumph in the democratization of diverse model systems. They allow for the economical sequencing of an entire genome or transcriptome of interest, and with technical variations can even provide insight into genome organization and the expression and regulation of genes. The analysis and biological interpretation of such large datasets can present significant challenges that depend on the 'scientific status' of the model system. While high-quality genome and transcriptome references are readily available for well-established model systems, the establishment of such references for an emerging model system often requires extensive resources such as finances, expertise and computation capabilities. The de novo assembly of a transcriptome represents an excellent entry point for genetic and molecular studies in emerging model systems as it can efficiently assess gene content while also serving as a reference for differential gene expression studies. However, the process of de novo transcriptome assembly is non-trivial, and as a rule must be empirically optimized for every dataset. For the researcher working with an emerging model system, and with little to no experience with assembling and quantifying short-read data from the Illumina platform, these processes can be daunting. In this guide we outline the major challenges faced when establishing a reference transcriptome de novo and we provide advice on how to approach such an endeavor. We describe the major experimental and bioinformatic steps, provide some broad recommendations and cautions for the newcomer to de novo transcriptome assembly and differential gene expression analyses. Moreover, we provide an initial selection of tools that can assist in the journey from raw short-read data to assembled transcriptome and lists of differentially expressed genes.

DifferentialRegulation: a Bayesian hierarchical approach to identify differentially regulated genes

Motivation

Although transcriptomics data is typically used to analyse mature spliced mRNA, recent attention has focused on jointly investigating spliced and unspliced (or precursor-) mRNA, which can be used to study gene regulation and changes in gene expression production. Nonetheless, most methods for spliced/unspliced inference (such as RNA velocity tools) focus on individual samples, and rarely allow comparisons between groups of samples (e.g., healthy vs. diseased). Furthermore, this kind of inference is challenging, because spliced and unspliced mRNA abundance is characterized by a high degree of quantification uncertainty, due to the prevalence of multi-mapping reads, i.e., reads compatible with multiple transcripts (or genes), and/or with both their spliced and unspliced versions.

Results

Here, we present DifferentialRegulation, a Bayesian hierarchical method to discover changes between experimental conditions with respect to the relative abundance of unspliced mRNA (over the total mRNA). We model the quantification uncertainty via a latent variable approach, where reads are allocated to their gene/transcript of origin, and to the respective splice version. We designed several benchmarks where our approach shows good performance, in terms of sensitivity and error control, versus state-of-the-art competitors. Importantly, our tool is flexible, and works with both bulk and single-cell RNA-sequencing data.

Availability and implementation

DifferentialRegulation is distributed as a Bioconductor R package.

Disregarding multimappers leads to biases in the functional assessment of NGS data

BACKGROUND

Standard ChIP-seq and RNA-seq processing pipelines typically disregard sequencing reads whose origin is ambiguous ("multimappers"). This usual practice has potentially important consequences for the functional interpretation of the data: genomic elements belonging to clusters composed of highly similar members are left unexplored.

RESULTS

In particular, disregarding multimappers leads to the underrepresentation in epigenetic studies of recently active transposable elements, such as AluYa5, L1HS and SVAs. Furthermore, this common strategy also has implications for transcriptomic analysis: members of repetitive gene families, such the ones including major histocompatibility complex (MHC) class I and II genes, are under-quantified.

CONCLUSION

Revealing inherent biases that permeate routine tasks such as functional enrichment analysis, our results underscore the urgency of broadly adopting multimapper-aware bioinformatic pipelines -currently restricted to specific contexts or communities- to ensure the reliability of genomic and transcriptomic studies.

N-of-one differential gene expression without control samples using a deep generative model

Differential analysis of bulk RNA-seq data often suffers from lack of good controls. Here, we present a generative model that replaces controls, trained solely on healthy tissues. The unsupervised model learns a low-dimensional representation and can identify the closest normal representation for a given disease sample. This enables control-free, single-sample differential expression analysis. In breast cancer, we demonstrate how our approach selects marker genes and outperforms a state-of-the-art method. Furthermore, significant genes identified by the model are enriched in driver genes across cancers. Our results show that the in silico closest normal provides a more favorable comparison than control samples.

A scoping review on deep learning for next-generation RNA-Seq. data analysis

In the last decade, transcriptome research adopting next-generation sequencing (NGS) technologies has gathered incredible momentum amongst functional genomics scientists, particularly amongst clinical/biomedical research groups. The progressive enfoldment/adoption of NGS technologies has incited an abundance of next-generation transcriptomic data harbouring an opulence of new knowledge in public databases. Nevertheless, knowledge discovery from these next-generation RNA-Seq. data analysis necessitates extensive bioinformatics know-how besides elaborate data analysis software packages consistent with the type and context of data analysis. Several reliability and reproducibility concerns continue to impede RNA-Seq. data analysis. Characteristic challenges comprise of data quality, hardware and networking provisions, selection and prioritisation of data analysis tools, and yet significantly implementing of robust machine learning algorithms for maximised exploitation of these experimental transcriptomic data. Over the years, numerous machine learning algorithms have been implemented for improved transcriptomic data analysis executing predominantly shallow learning approaches. More recently, deep learning algorithms are becoming more mainstream, and enactment for next-generation RNA-Seq. data analysis could be revolutionary in the coming years in the biomedical domain. In this scoping review, we attempt to determine the existing literature's size and potential nature in deep learning and NGS RNA-Seq. data analysis. An analysis of the contemporary topics of next-generation RNA-Seq. data analysis based on deep learning algorithms is critically reviewed, emphasising open-source resources.

Analyzing RNA-Seq Gene Expression Data Using Deep Learning Approaches for Cancer Classification

Ribonucleic acid Sequencing (RNA-Seq) analysis is particularly useful for obtaining insights into differentially expressed genes. However, it is challenging because of its high-dimensional data. Such analysis is a tool with which to find underlying patterns in data, e.g., for cancer specific biomarkers. In the past, analyses were performed on RNA-Seq data pertaining to the same cancer class as positive and negative samples, i.e., without samples of other cancer types. To perform multiple cancer type classification and to find differentially expressed genes, data for multiple cancer types need to be analyzed. Several repositories offer RNA-Seq data for various cancer types. In this paper, data from the Mendeley data repository for five cancer types are analyzed. As a first step, RNA-Seq values are converted to 2D images using normalization and zero padding. In the next step, relevant features are extracted and selected using Deep Learning (DL). In the last phase, classification is performed, and eight DL algorithms are used. Results and discussion are based on four different splitting strategies and k-fold cross validation for each DL classifier. Furthermore, a comparative analysis is performed with state of the art techniques discussed in literature. The results demonstrated that classifiers performed best at 70–30 split, and that Convolutional Neural Network (CNN) achieved the best overall results. Hence, CNN is the best DL model for classification among the eight studied DL models, and is easy to implement and simple to understand.

MGcount: a total RNA-seq quantification tool to address multi-mapping and multi-overlapping alignments ambiguity in non-coding transcripts

BACKGROUND

Total-RNA sequencing (total-RNA-seq) allows the simultaneous study of both the coding and the non-coding transcriptome. Yet, computational pipelines have traditionally focused on particular biotypes, making assumptions that are not fullfilled by total-RNA-seq datasets. Transcripts from distinct RNA biotypes vary in length, biogenesis, and function, can overlap in a genomic region, and may be present in the genome with a high copy number. Consequently, reads from total-RNA-seq libraries may cause ambiguous genomic alignments, demanding for flexible quantification approaches.

RESULTS

Here we present Multi-Graph count (MGcount), a total-RNA-seq quantification tool combining two strategies for handling ambiguous alignments. First, MGcount assigns reads hierarchically to small-RNA and long-RNA features to account for length disparity when transcripts overlap in the same genomic position. Next, MGcount aggregates RNA products with similar sequences where reads systematically multi-map using a graph-based approach. MGcount outputs a transcriptomic count matrix compatible with RNA-sequencing downstream analysis pipelines, with both bulk and single-cell resolution, and the graphs that model repeated transcript structures for different biotypes. The software can be used as a python module or as a single-file executable program.

CONCLUSIONS

MGcount is a flexible total-RNA-seq quantification tool that successfully integrates reads that align to multiple genomic locations or that overlap with multiple gene features. Its approach is suitable for the simultaneous estimation of protein-coding, long non-coding and small non-coding transcript concentration, in both precursor and processed forms. Both source code and compiled software are available at https://github.com/hitaandrea/MGcount .

A method for simultaneous detection of small and long RNA biotypes by ribodepleted RNA-Seq

RNA sequencing offers unprecedented access to the transcriptome. Key to this is the identification and quantification of many different species of RNA from the same sample at the same time. In this study we describe a novel protocol for simultaneous detection of coding and non-coding transcripts using modifications to the Ion Total RNA-Seq kit v2 protocol, with integration of QIASeq FastSelect rRNA removal kit. We report highly consistent sequencing libraries can be produced from both frozen high integrity mouse hippocampal tissue and the more challenging post-mortem human tissue. Removal of rRNA using FastSelect was extremely efficient, resulting in less than 1.5% rRNA content in the final library. We identified > 30,000 unique transcripts from all samples, including protein-coding genes and many species of non-coding RNA, in biologically-relevant proportions. Furthermore, the normalized sequencing read count for select genes significantly negatively correlated with Ct values from qRT-PCR analysis from the same samples. These results indicate that this protocol accurately and consistently identifies and quantifies a wide variety of transcripts simultaneously. The highly efficient rRNA depletion, coupled with minimized sample handling and without complicated and high-loss size selection protocols, makes this protocol useful to researchers wishing to investigate whole transcriptomes.

Unsupervised Machine Learning-Based Analysis of Clinical Features, Bone Mineral Density Features and Medical Care Costs of Rotator Cuff Tears

Purpose

We aim to present unsupervised machine learning-based analysis of clinical features, bone mineral density (BMD) features, and medical care costs of Rotator cuff tears (RCT).

Patients and Methods

Fifty-three patients with RCT were reviewed, the clinical features, BMD features, and medical care costs were collected and analyzed by descriptive statistics. Furtherly, unsupervised machine learning (UML) algorithm was used for dimensionality reduction and cluster analysis of the RCT data.

Results

There were 26 males and 27 females. The patients were divided into four subgroups using the UML algorithm. There were significant differences among four subgroups regarding trauma exposure, full-thickness supraspinatus tendon tears, infraspinatus tendon tear, subscapularis tendon tear, BMD distribution, medial row anchors, lateral row anchors, total medical care costs, and consumables costs. We observed the highest frequency of trauma exposure, infraspinatus tendon tear, subscapularis tendon tear, osteoporosis, the highest number of medial row anchors, lateral row anchors, total medical care costs, and consumables costs in subgroup II.

Conclusion

The unsupervised machine learning-based analysis of RCT can provide clinically meaningful classification, which shows good interpretability and contribute to a better understanding of RCT. The significance of the results is limited due to the small number of samples, a larger follow-up study is needed to confirm the encouraging results.

Non-coding Natural Antisense Transcripts: Analysis and Application

Non-coding natural antisense transcripts (ncNATs) are regulatory RNA sequences that are transcribed in the opposite direction to protein-coding or non-coding transcripts. These transcripts are implicated in a broad variety of biological and pathological processes, including tumorigenesis and oncogenic progression. With this complex field still in its infancy, annotations, expression profiling and functional characterisations of ncNATs are far less comprehensive than those for protein-coding genes, pointing out substantial gaps in the analysis and characterisation of these regulatory transcripts. In this review, we discuss ncNATs from an analysis perspective, in particular regarding the use of high-throughput sequencing strategies, such as RNA-sequencing, and summarize the unique challenges of investigating the antisense transcriptome. Finally, we elaborate on their potential as biomarkers and future targets for treatment, focusing on cancer.

Integrative Methods and Practical Challenges for Single-Cell Multi-omics

Fast-developing single-cell multimodal omics (scMulti-omics) technologies enable the measurement of multiple modalities, such as DNA methylation, chromatin accessibility, RNA expression, protein abundance, gene perturbation, and spatial information, from the same cell. scMulti-omics can comprehensively explore and identify cell characteristics, while also presenting challenges to the development of computational methods and tools for integrative analyses. Here, we review these integrative methods and summarize the existing tools for studying a variety of scMulti-omics data. The various functionalities and practical challenges in using the available tools in the public domain are explored through several case studies. Finally, we identify remaining challenges and future trends in scMulti-omics modeling and analyses.

mAHTPred: a sequence-based meta-predictor for improving the prediction of anti-hypertensive peptides using effective feature representation

MOTIVATION

Cardiovascular disease is the primary cause of death globally accounting for approximately 17.7 million deaths per year. One of the stakes linked with cardiovascular diseases and other complications is hypertension. Naturally derived bioactive peptides with antihypertensive activities serve as promising alternatives to pharmaceutical drugs. So far, there is no comprehensive analysis, assessment of diverse features and implementation of various machine-learning (ML) algorithms applied for antihypertensive peptide (AHTP) model construction.

RESULTS

In this study, we utilized six different ML algorithms, namely, Adaboost, extremely randomized tree (ERT), gradient boosting (GB), k-nearest neighbor, random forest (RF) and support vector machine (SVM) using 51 feature descriptors derived from eight different feature encodings for the prediction of AHTPs. While ERT-based trained models performed consistently better than other algorithms regardless of various feature descriptors, we treated them as baseline predictors, whose predicted probability of AHTPs was further used as input features separately for four different ML-algorithms (ERT, GB, RF and SVM) and developed their corresponding meta-predictors using a two-step feature selection protocol. Subsequently, the integration of four meta-predictors through an ensemble learning approach improved the balanced prediction performance and model robustness on the independent dataset. Upon comparison with existing methods, mAHTPred showed superior performance with an overall improvement of approximately 6-7% in both benchmarking and independent datasets.

AVAILABILITY AND IMPLEMENTATION

The user-friendly online prediction tool, mAHTPred is freely accessible at http://thegleelab.org/mAHTPred.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Handling multi-mapped reads in RNA-seq

Many eukaryotic genomes harbour large numbers of duplicated sequences, of diverse biotypes, resulting from several mechanisms including recombination, whole genome duplication and retro-transposition. Such repeated sequences complicate gene/transcript quantification during RNA-seq analysis due to reads mapping to more than one locus, sometimes involving genes embedded in other genes. Genes of different biotypes have dissimilar levels of sequence duplication, with long-noncoding RNAs and messenger RNAs sharing less sequence similarity to other genes than biotypes encoding shorter RNAs. Many strategies have been elaborated to handle these multi-mapped reads, resulting in increased accuracy in gene/transcript quantification, although separate tools are typically used to estimate the abundance of short and long genes due to their dissimilar characteristics. This review discusses the mechanisms leading to sequence duplication, the biotypes affected, the computational strategies employed to deal with multi-mapped reads and the challenges that still remain to be overcome.

NCResNet: Noncoding Ribonucleic Acid Prediction Based on a Deep Resident Network of Ribonucleic Acid Sequences

Noncoding RNA (ncRNA) is a kind of RNA that plays an important role in many biological processes, diseases, and cancers, while cannot translate into proteins. With the development of next-generation sequence technology, thousands of novel RNAs with long open reading frames (ORFs, longest ORF length > 303 nt) and short ORFs (longest ORF length ≤ 303 nt) have been discovered in a short time. How to identify ncRNAs more precisely from novel unannotated RNAs is an important step for RNA functional analysis, RNA regulation, etc. However, most previous methods only utilize the information of sequence features. Meanwhile, most of them have focused on long-ORF RNA sequences, but not adapted to short-ORF RNA sequences. In this paper, we propose a new reliable method called NCResNet. NCResNet employs 57 hybrid features of four categories as inputs, including sequence, protein, RNA structure, and RNA physicochemical properties, and introduces feature enhancement and deep feature learning policies in a neural net model to adapt to this problem. The experiments on benchmark datasets of 8 species shows NCResNet has higher accuracy and higher Matthews correlation coefficient (MCC) compared with other state-of-the-art methods. Particularly, on four short-ORF RNA sequence datasets, specifically mouse, Saccharomyces cerevisiae, zebrafish, and cow, NCResNet achieves greater than 10 and 15% improvements over other state-of-the-art methods in terms of accuracy and MCC. Meanwhile, for long-ORF RNA sequence datasets, NCResNet also has better accuracy and MCC than other state-of-the-art methods on most test datasets. Codes and data are available at https://github.com/abcair/NCResNet.

Single-Cell RNA Sequencing of Plant-Associated Bacterial Communities

Plants in soil are not solitary, hence continually interact with and obtain benefits from a community of microbes ("microbiome"). The meta-functional output from the microbiome results from complex interactions among the different community members with distinct taxonomic identities and metabolic capacities. Particularly, the bacterial communities of the root surface are spatially organized structures composed of root-attached biofilms and planktonic cells arranged in complex layers. With the distinct but coordinated roles among the different member cells, bacterial communities resemble properties of a multicellular organism. High throughput sequencing technologies have allowed rapid and large-scale analysis of taxonomic composition and metabolic capacities of bacterial communities. However, these methods are generally unable to reconstruct the assembly of these communities, or how the gene expression patterns in individual cells/species are coordinated within these communities. Single-cell transcriptomes of community members can identify how gene expression patterns vary among members of the community, including differences among different cells of the same species. This information can be used to classify cells based on functional gene expression patterns, and predict the spatial organization of the community. Here we discuss strategies for the isolation of single bacterial cells, mRNA enrichment, library construction, and analysis and interpretation of the resulting single-cell RNA-Seq datasets. Unraveling regulatory and metabolic processes at the single cell level is expected to yield an unprecedented discovery of mechanisms involved in bacterial recruitment, attachment, assembly, organization of the community, or in the specific interactions among the different members of these communities.