Third-generation sequencing technologies benefit transcriptome analysis by generating longer sequencing reads. However, not all single-molecule long reads represent full transcripts due to incomplete cDNA synthesis and the sequencing length limit of the platform. This drives a need for long read transcript assembly. We quantify the benefit that can be achieved by using a transcript assembler on long reads. Adding long-read-specific algorithms, we evolved Scallop to make Scallop-LR, a long-read transcript assembler, to handle the computational challenges arising from long read lengths and high error rates. Analyzing 26 SRA PacBio datasets using Scallop-LR, Iso-Seq Analysis, and StringTie, we quantified the amount by which assembly improved Iso-Seq results. Through combined evaluation methods, we found that Scallop-LR identifies 2100–4000 more (for 18 human datasets) or 1100–2200 more (for eight mouse datasets) known transcripts than Iso-Seq Analysis, which does not do assembly. Further, Scallop-LR finds 2.4–4.4 times more potentially novel isoforms than Iso-Seq Analysis for the human and mouse datasets. StringTie also identifies more transcripts than Iso-Seq Analysis. Adding long-read-specific optimizations in Scallop-LR increases the numbers of predicted known transcripts and potentially novel isoforms for the human transcriptome compared to several recent short-read assemblers (e.g. StringTie). Our findings indicate that transcript assembly by Scallop-LR can reveal a more complete human transcriptome.
Haplotype phasing of genetic variants is important for interpretation of the maize genome, population genetic analysis, and functional genomic analysis of allelic activity. Accordingly, accurate methods for phasing full-length isoforms are essential for functional genomics study. In this study, we performed an isoform-level phasing study in maize, using two inbred lines and their reciprocal crosses, based on single-molecule full-length cDNA sequencing. To phase and analyze full-length transcripts between hybrids and parents, we developed a tool called IsoPhase. Using this tool, we validated the majority of SNPs called against matching short read data and identified cases of allele-specific, gene-level, and isoform-level expression. Our results revealed that maize parental and hybrid lines exhibit different splicing activities. After phasing 6,847 genes in two reciprocal hybrids using embryo, endosperm and root tissues, we annotated the SNPs and identified large-effect genes. In addition, based on single-molecule sequencing, we identified parent-of-origin isoforms in maize hybrids, different novel isoforms between maize parent and hybrid lines, and imprinted genes from different tissues. Finally, we characterized variation in cis- and trans-regulatory effects. Our study provides measures of haplotypic expression that could increase power and accuracy in studies of allelic expression.
Long-read next-generation amplicon sequencing shows promise for studying complete genes or genomes from complex and diverse populations. Current long-read sequencing technologies have challenging error profiles, hindering data processing and incorporation into downstream analyses. Here we consider the problem of how to reconstruct, free of sequencing error, the true sequence variants and their associated frequencies from PacBio reads. Called ‘amplicon denoising’, this problem has been extensively studied for short-read sequencing technologies, but current solutions do not always successfully generalize to long reads with high indel error rates. We introduce two methods: one that runs nearly instantly and is very accurate for medium length reads and high template coverage, and another, slower method that is more robust when reads are very long or coverage is lower. On two Mock Virus Community datasets with ground truth, each sequenced on a different PacBio instrument, and on a number of simulated datasets, we compare our two approaches to each other and to existing algorithms. We outperform all tested methods in accuracy, with competitive run times even for our slower method, successfully discriminating templates that differ by a just single nucleotide. Julia implementations of Fast Amplicon Denoising (FAD) and Robust Amplicon Denoising (RAD), and a webserver interface, are freely available. © The Author(s) 2019. Published by Oxford University Press on behalf of Nucleic Acids Research.
In the past several years, single-molecule sequencing platforms, such as those by Pacific Biosciences and Oxford Nanopore Technologies, have become available to researchers and are currently being tested for clinical applications. They offer exceptionally long reads that permit direct sequencing through regions of the genome inaccessible or difficult to analyze by short-read platforms. This includes disease-causing long repetitive elements, extreme GC content regions, and complex gene loci. Similarly, these platforms enable structural variation characterization at previously unparalleled resolution and direct detection of epigenetic marks in native DNA. Here, we review how these technologies are opening up new clinical avenues that are being applied to pathogenic microorganisms and viruses, constitutional disorders, pharmacogenomics, cancer, and more.Copyright © 2018 Elsevier Ltd. All rights reserved.
TranscriptClean: variant-aware correction of indels, mismatches and splice junctions in long-read transcripts.
Long-read, single-molecule sequencing platforms hold great potential for isoform discovery and characterization of multi-exon transcripts. However, their high error rates are an obstacle to distinguishing novel transcript isoforms from sequencing artifacts. Therefore, we developed the package TranscriptClean to correct mismatches, microindels and noncanonical splice junctions in mapped transcripts using the reference genome while preserving known variants.Our method corrects nearly all mismatches and indels present in a publically available human PacBio Iso-seq dataset, and rescues 39% of noncanonical splice junctions.All Python and R scripts used in this paper are available at https://github.com/dewyman/TranscriptClean.
Hybrid sequencing-based personal full-length transcriptomic analysis implicates proteostatic stress in metastatic ovarian cancer.
Comprehensive molecular characterization of myriad somatic alterations and aberrant gene expressions at personal level is key to precision cancer therapy, yet limited by current short-read sequencing technology, individualized catalog of complete genomic and transcriptomic features is thus far elusive. Here, we integrated second- and third-generation sequencing platforms to generate a multidimensional dataset on a patient affected by metastatic epithelial ovarian cancer. Whole-genome and hybrid transcriptome dissection captured global genetic and transcriptional variants at previously unparalleled resolution. Particularly, single-molecule mRNA sequencing identified a vast array of unannotated transcripts, novel long noncoding RNAs and gene chimeras, permitting accurate determination of transcription start, splice, polyadenylation and fusion sites. Phylogenetic and enrichment inference of isoform-level measurements implicated early functional divergence and cytosolic proteostatic stress in shaping ovarian tumorigenesis. A complementary imaging-based high-throughput drug screen was performed and subsequently validated, which consistently pinpointed proteasome inhibitors as an effective therapeutic regime by inducing protein aggregates in ovarian cancer cells. Therefore, our study suggests that clinical application of the emerging long-read full-length analysis for improving molecular diagnostics is feasible and informative. An in-depth understanding of the tumor transcriptome complexity allowed by leveraging the hybrid sequencing approach lays the basis to reveal novel and valid therapeutic vulnerabilities in advanced ovarian malignancies.
PacBio full-length cDNA sequencing integrated with RNA-seq reads drastically improves the discovery of splicing transcripts in rice.
In eukaryotes, alternative splicing (AS) greatly expands the diversity of transcripts. However, it is challenging to accurately determine full-length splicing isoforms. Recently, more studies have taken advantage of Pacific Bioscience (PacBio) long-read sequencing to identify full-length transcripts. Nevertheless, the high error rate of PacBio reads seriously offsets the advantages of long reads, especially for accurately identifying splicing junctions. To best capitalize on the features of long reads, we used Illumina RNA-seq reads to improve PacBio circular consensus sequence (CCS) quality and to validate splicing patterns in the rice transcriptome. We evaluated the impact of CCS accuracy on the number and the validation rate of splicing isoforms, and integrated a comprehensive pipeline of splicing transcripts analysis by Iso-Seq and RNA-seq (STAIR) to identify the full-length multi-exon isoforms in rice seedling transcriptome (Oryza sativa L. ssp. japonica). STAIR discovered 11 733 full-length multi-exon isoforms, 6599 more than the SMRT Portal RS_IsoSeq pipeline did. Of these splicing isoforms identified, 4453 (37.9%) were missed in assembled transcripts from RNA-seq reads, and 5204 (44.4%), including 268 multi-exon long non-coding RNAs (lncRNAs), were not reported in the MSU_osa1r7 annotation. Some randomly selected unreported splicing junctions were verified by polymerase chain reaction (PCR) amplification. In addition, we investigated alternative polyadenylation (APA) events in transcripts and identified 829 major polyadenylation [poly(A)] site clusters (PACs). The analysis of splicing isoforms and APA events will facilitate the annotation of the rice genome and studies on the expression and polyadenylation of AS genes in different developmental stages or growth conditions of rice. © 2018 The Authors The Plant Journal © 2018 John Wiley & Sons Ltd.
Genome sequencing and CRISPR/Cas9 gene editing of an early flowering Mini-Citrus (Fortunella hindsii).
Hongkong kumquat (Fortunella hindsii) is a wild citrus species characterized by dwarf plant height and early flowering. Here, we identified the monoembryonic F. hindsii (designated as ‘Mini-Citrus’) for the first time and constructed its selfing lines. This germplasm constitutes an ideal model for the genetic and functional genomics studies of citrus, which have been severely hindered by the long juvenility and inherent apomixes of citrus. F. hindsii showed a very short juvenile period (~8 months) and stable monoembryonic phenotype under cultivation. We report the first de novo assembled 373.6 Mb genome sequences (Contig-N50 2.2 Mb and Scaffold-N50 5.2 Mb) for F. hindsii. In total, 32 257 protein-coding genes were annotated, 96.9% of which had homologues in other eight Citrinae species. The phylogenomic analysis revealed a close relationship of F. hindsii with cultivated citrus varieties, especially with mandarin. Furthermore, the CRISPR/Cas9 system was demonstrated to be an efficient strategy to generate target mutagenesis on F. hindsii. The modifications of target genes in the CRISPR-modified F. hindsii were predominantly 1-bp insertions or small deletions. This genetic transformation system based on F. hindsii could shorten the whole process from explant to T1 mutant to about 15 months. Overall, due to its short juvenility, monoembryony, close genetic background to cultivated citrus and applicability of CRISPR, F. hindsii shows unprecedented potentials to be used as a model species for citrus research. © 2019 The Authors. Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd.
The Populus shoot undergoes primary growth (longitudinal growth) followed by secondary growth (radial growth), which produces biomass that is an important source of energy worldwide. We adopted joint PacBio Iso-Seq and RNA-seq analysis to identify differentially expressed transcripts along a developmental gradient from the shoot apex to the fifth internode of Populus Nanlin895. We obtained 87 150 full-length transcripts, including 2081 new isoforms and 62 058 new alternatively spliced isoforms, most of which were produced by intron retention, that were used to update the Populus annotation. Among these novel isoforms, there are 1187 long non-coding RNAs and 356 fusion genes. Using this annotation, we found 15 838 differentially expressed transcripts along the shoot developmental gradient, of which 1216 were transcription factors (TFs). Only a few of these genes were reported previously. The differential expression of these TFs suggests that they may play important roles in primary and secondary growth. AP2, ARF, YABBY and GRF TFs are highly expressed in the apex, whereas NAC, bZIP, PLATZ and HSF TFs are likely to be important for secondary growth. Overall, our findings provide evidence that long-read sequencing can complement short-read sequencing for cataloguing and quantifying eukaryotic transcripts and increase our understanding of the vital and dynamic process of shoot development. © 2018 The Authors. Plant Biotechnology Journal published by Society for Experimental Biology and The Association of Applied Biologists and John Wiley & Sons Ltd.
Genome and transcriptome sequencing of the astaxanthin-producing green microalga, Haematococcus pluvialis.
Haematococcus pluvialis is a freshwater species of Chlorophyta, family Haematococcaceae. It is well known for its capacity to synthesize high amounts of astaxanthin, which is a strong antioxidant that has been utilized in aquaculture and cosmetics. To improve astaxanthin yield and to establish genetic resources for H. pluvialis, we performed whole-genome sequencing, assembly, and annotation of this green microalga. A total of 83.1 Gb of raw reads were sequenced. After filtering the raw reads, we subsequently generated a draft assembly with a genome size of 669.0?Mb, a scaffold N50 of 288.6?kb, and predicted 18,545 genes. We also established a robust phylogenetic tree from 14 representative algae species. With additional transcriptome data, we revealed some novel potential genes that are involved in the synthesis, accumulation, and regulation of astaxanthin production. In addition, we generated an isoform-level reference transcriptome set of 18,483 transcripts with high confidence. Alternative splicing analysis demonstrated that intron retention is the most frequent mode. In summary, we report the first draft genome of H. pluvialis. These genomic resources along with transcriptomic data provide a solid foundation for the discovery of the genetic basis for theoretical and commercial astaxanthin enrichment.
Insights into the evolution and drug susceptibility of Babesia duncani from the sequence of its mitochondrial and apicoplast genomes.
Babesia microti and Babesia duncani are the main causative agents of human babesiosis in the United States. While significant knowledge about B. microti has been gained over the past few years, nothing is known about B. duncani biology, pathogenesis, mode of transmission or sensitivity to currently recommended therapies. Studies in immunocompetent wild type mice and hamsters have shown that unlike B. microti, infection with B. duncani results in severe pathology and ultimately death. The parasite factors involved in B. duncani virulence remain unknown. Here we report the first known completed sequence and annotation of the apicoplast and mitochondrial genomes of B. duncani. We found that the apicoplast genome of this parasite consists of a 34?kb monocistronic circular molecule encoding functions that are important for apicoplast gene transcription as well as translation and maturation of the organelle’s proteins. The mitochondrial genome of B. duncani consists of a 5.9?kb monocistronic linear molecule with two inverted repeats of 48?bp at both ends. Using the conserved cytochrome b (Cytb) and cytochrome c oxidase subunit I (coxI) proteins encoded by the mitochondrial genome, phylogenetic analysis revealed that B. duncani defines a new lineage among apicomplexan parasites distinct from B. microti, Babesia bovis, Theileria spp. and Plasmodium spp. Annotation of the apicoplast and mitochondrial genomes of B. duncani identified targets for development of effective therapies. Our studies set the stage for evaluation of the efficacy of these drugs alone or in combination against B. duncani in culture as well as in animal models.Copyright © 2018 Australian Society for Parasitology. Published by Elsevier Ltd. All rights reserved.
A global survey of full-length transcriptome of Ginkgo biloba reveals transcript variants involved in flavonoid biosynthesis
Ginkgo biloba, which contains flavonoids as bioactive components, is widely used in traditional Chinese medicine. Increasing the flavonoid production of medicinal plants through genetic engineering generally focuses on the key genes involved in flavonoid biosynthesis. However, the molecular mechanisms underlying such biosynthesis are not yet well understood. To understand these mechanisms, a combination of second-generation sequencing (SGS) and single-molecule real-time (SMRT) sequencing was applied to G. biloba. Eight tissues were sampled for SMRT sequencing to generate a high-quality, full-length transcriptome database. From 23.36 Gb clean reads, 12,954 alternative polyadenylation events, 12,290 alternative splicing events, 929 fusion transcripts, 2,286 novel transcripts, and 1,270 lncRNAs were predicted by removing redundant reads. Further studies reveal that 7 AS, 5 lncRNA, and 6 fusion gene events were identified in flavonoid biosynthesis. A total of 12 gene modules were revealed to be involved in flavonoid metabolism structural genes and transcription factors by constructing co-expression networks. Weighted gene coexpression network analysis (WGCNA) analysis reveals that some hub genes operate during the biosynthesis by identifying transcription factors (TFs) and structure genes. Seven key hub genes were also identified by analyzing the correlation between gene expression level and flavonoids content. The results highlight the importance of SMRT sequencing of the full-length transcriptome in improving genome annotation and elucidating the gene regulation of flavonoid biosynthesis in G. biloba by providing a comprehensive set of reference transcripts.
Platanus-allee is a de novo haplotype assembler enabling a comprehensive access to divergent heterozygous regions.
The ultimate goal for diploid genome determination is to completely decode homologous chromosomes independently, and several phasing programs from consensus sequences have been developed. These methods work well for lowly heterozygous genomes, but the manifold species have high heterozygosity. Additionally, there are highly divergent regions (HDRs), where the haplotype sequences differ considerably. Because HDRs are likely to direct various interesting biological phenomena, many genomic analysis targets fall within these regions. However, they cannot be accessed by existing phasing methods, and we have to adopt costly traditional methods. Here, we develop a de novo haplotype assembler, Platanus-allee ( http://platanus.bio.titech.ac.jp/platanus2 ), which initially constructs each haplotype sequence and then untangles the assembly graphs utilizing sequence links and synteny information. A comprehensive benchmark analysis reveals that Platanus-allee exhibits high recall and precision, particularly for HDRs. Using this approach, previously unknown HDRs are detected in the human genome, which may uncover novel aspects of genome variability.
Normalization of cDNA is widely used to improve the coverage of rare transcripts in analysis of transcriptomes employing next-generation sequencing. Recently, long-read technology has been emerging as a powerful tool for sequencing and construction of transcriptomes, especially for complex genomes containing highly similar transcripts and transcript-spliced isoforms. Here, we analyzed the transcriptome of sugarcane, with a highly polyploidy plant genome, by PacBio isoform sequencing (Iso-Seq) of two different cDNA library preparations, with and without a normalization step. The results demonstrated that, while the two libraries included many of the same transcripts, many longer transcripts were removed and many new generally shorter transcripts were detected by normalization. For the same input cDNA and the same data yield, the normalized library recovered more total transcript isoforms, number of predicted gene families and orthologous groups, resulting in a higher representation for the sugarcane transcriptome, compared to the non-normalized library. The non-normalized library, on the other hand, included a wider transcript length range with more longer transcripts above ~1.25 kb, more transcript isoforms per gene family and gene ontology terms per transcript. A large proportion of the unique transcripts comprising ~52% of the normalized library were expressed at a lower level than the unique transcripts from the non-normalized library, across three tissue types tested including leaf, stalk and root. About 83% of the total 5,348 predicted long noncoding transcripts was derived from the normalized library, of which ~80% was derived from the lowly expressed fraction. Functional annotation of the unique transcripts suggested that each library enriched different functional transcript fractions. This demonstrated the complementation of the two approaches in obtaining a complete transcriptome of a complex genome at the sequencing depth used in this study.
Reconstruction of the full-length transcriptome atlas using PacBio Iso-Seq provides insight into the alternative splicing in Gossypium australe.
Gossypium australe F. Mueller (2n?=?2x?=?26, G2 genome) possesses valuable characteristics. For example, the delayed gland morphogenesis trait causes cottonseed protein and oil to be edible while retaining resistance to biotic stress. However, the lack of gene sequences and their alternative splicing (AS) in G. australe remain unclear, hindering to explore species-specific biological morphogenesis.Here, we report the first sequencing of the full-length transcriptome of the Australian wild cotton species, G. australe, using Pacific Biosciences single-molecule long-read isoform sequencing (Iso-Seq) from the pooled cDNA of ten tissues to identify transcript loci and splice isoforms. We reconstructed the G. australe full-length transcriptome and identified 25,246 genes, 86 pre-miRNAs and 1468 lncRNAs. Most genes (12,832, 50.83%) exhibited two or more isoforms, suggesting a high degree of transcriptome complexity in G. australe. A total of 31,448 AS events in five major types were found among the 9944 gene loci. Among these five major types, intron retention was the most frequent, accounting for 68.85% of AS events. 29,718 polyadenylation sites were detected from 14,536 genes, 7900 of which have alternative polyadenylation sites (APA). In addition, based on our AS events annotations, RNA-Seq short reads from germinating seeds showed that differential expression of these events occurred during seed germination. Ten AS events that were randomly selected were further confirmed by RT-PCR amplification in leaf and germinating seeds.The reconstructed gene sequences and their AS in G. australe would provide information for exploring beneficial characteristics in G. australe.