Sequence deeper without sequencing more: Bayesian resolution of ambiguously mapped reads

Next-generation sequencing (NGS) has transformed molecular biology and contributed to many seminal insights into genomic regulation and function. Apart from whole-genome sequencing, an NGS workflow involves alignment of the sequencing reads to the genome of study, after which the resulting alignments can be used for downstream analyses. However, alignment is complicated by the repetitive sequences; many reads align to more than one genomic locus, with 15–30% of the genome not being uniquely mappable by short-read NGS. This problem is typically addressed by discarding reads that do not uniquely map to the genome, but this practice can lead to systematic distortion of the data. Previous studies that developed methods for handling ambiguously mapped reads were often of limited applicability or were computationally intensive, hindering their broader usage. In this work, we present SmartMap: an algorithm that augments industry-standard aligners to enable usage of ambiguously mapped reads by assigning weights to each alignment with Bayesian analysis of the read distribution and alignment quality. SmartMap is computationally efficient, utilizing far fewer weighting iterations than previously thought necessary to process alignments and, as such, analyzing more than a billion alignments of NGS reads in approximately one hour on a desktop PC. By applying SmartMap to peak-type NGS data, including MNase-seq, ChIP-seq, and ATAC-seq in three organisms, we can increase read depth by up to 53% and increase the mapped proportion of the genome by up to 18% compared to analyses utilizing only uniquely mapped reads. We further show that SmartMap enables the analysis of more than 140,000 repetitive elements that could not be analyzed by traditional ChIP-seq workflows, and we utilize this method to gain insight into the epigenetic regulation of different classes of repetitive elements. These data emphasize both the dangers of discarding ambiguously mapped reads and their power for driving biological discovery.     

JunctionViewer: customizable annotation software for repeat-rich genomic regions

Background  

Repeat-rich regions such as centromeres receive less attention than their gene-rich euchromatic counterparts because the former are difficult to assemble and analyze. Our objectives were to 1) map all ten centromeres onto the maize genetic map and 2) characterize the sequence features of maize centromeres, each of which spans several megabases of highly repetitive DNA. Repetitive sequences can be mapped using special molecular markers that are based on PCR with primers designed from two unique "repeat junctions". Efficient screening of large amounts of maize genome sequence data for repeat junctions, as well as key centromere sequence features required the development of specific annotation software.     

Mobile element insertions are frequent in oesophageal adenocarcinomas and can mislead paired-end sequencing analysis

Background

Mobile elements are active in the human genome, both in the germline and cancers, where they can mutate driver genes.

Results

While analysing whole genome paired-end sequencing of oesophageal adenocarcinomas to find genomic rearrangements, we identified three ways in which new mobile element insertions appear in the data, resembling translocation or insertion junctions: inserts where unique sequence has been transduced by an L1 (Long interspersed element 1) mobile element; novel inserts that are confidently, but often incorrectly, mapped by alignment software to L1s or polyA tracts in the reference sequence; and a combination of these two ways, where different sequences within one insert are mapped to different loci. We identified nine unique sequences that were transduced by neighbouring L1s, both L1s in the reference genome and L1s not present in the reference. Many of the resulting inserts were small fragments that include little or no recognisable mobile element sequence. We found 6 loci in the reference genome to which sequence reads from inserts were frequently mapped, probably erroneously, by alignment software: these were either L1 sequence or particularly long polyA runs. Inserts identified from such apparent rearrangement junctions averaged 16 inserts/tumour, range 0–153 insertions in 43 tumours. However, many inserts would not be detected by mapping the sequences to the reference genome, because they do not include sufficient mappable sequence. To estimate total somatic inserts we searched for polyA sequences that were not present in the matched normal or other normals from the same tumour batch, and were not associated with known polymorphisms. Samples of these candidate inserts were verified by sequencing across them or manual inspection of surrounding reads: at least 85 % were somatic and resembled L1-mediated events, most including L1Hs sequence. Approximately 100 such inserts were detected per tumour on average (range zero to approximately 700).

Conclusions

Somatic mobile elements insertions are abundant in these tumours, with over 75 % of cases having a number of novel inserts detected. The inserts create a variety of problems for the interpretation of paired-end sequencing data.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1685-z) contains supplementary material, which is available to authorized users.     

Mapping Accuracy of Short Reads from Massively Parallel Sequencing and the Implications for Quantitative Expression Profiling

Background

Massively parallel sequencing offers an enormous potential for expression profiling, in particular for interspecific comparisons. Currently, different platforms for massively parallel sequencing are available, which differ in read length and sequencing costs. The 454-technology offers the highest read length. The other sequencing technologies are more cost effective, on the expense of shorter reads. Reliable expression profiling by massively parallel sequencing depends crucially on the accuracy to which the reads could be mapped to the corresponding genes.

Methodology/Principal Findings

We performed an in silico analysis to evaluate whether incorrect mapping of the sequence reads results in a biased expression pattern. A comparison of six available mapping software tools indicated a considerable heterogeneity in mapping speed and accuracy. Independently of the software used to map the reads, we found that for compact genomes both short (35 bp, 50 bp) and long sequence reads (100 bp) result in an almost unbiased expression pattern. In contrast, for species with a larger genome containing more gene families and repetitive DNA, shorter reads (35–50 bp) produced a considerable bias in gene expression. In humans, about 10% of the genes had fewer than 50% of the sequence reads correctly mapped. Sequence polymorphism up to 9% had almost no effect on the mapping accuracy of 100 bp reads. For 35 bp reads up to 3% sequence divergence did not affect the mapping accuracy strongly. The effect of indels on the mapping efficiency strongly depends on the mapping software.

Conclusions/Significance

In complex genomes, expression profiling by massively parallel sequencing could introduce a considerable bias due to incorrectly mapped sequence reads if the read length is short. Nevertheless, this bias could be accounted for if the genomic sequence is known. Furthermore, sequence polymorphisms and indels also affect the mapping accuracy and may cause a biased gene expression measurement. The choice of the mapping software is highly critical and the reliability depends on the presence/absence of indels and the divergence between reads and the reference genome. Overall, we found SSAHA2 and CLC to produce the most reliable mapping results.     

iMapper: a web application for the automated analysis and mapping of insertional mutagenesis sequence data against Ensembl genomes

SUMMARY: Insertional mutagenesis is a powerful method for gene discovery. To identify the location of insertion sites in the genome linker based polymerase chain reaction (PCR) methods (such as splinkerette-PCR) may be employed. We have developed a web application called iMapper (Insertional Mutagenesis Mapping and Analysis Tool) for the efficient analysis of insertion site sequence reads against vertebrate and invertebrate Ensembl genomes. Taking linker based sequences as input, iMapper scans and trims the sequence to remove the linker and sequences derived from the insertional mutagen. The software then identifies and removes contaminating sequences derived from chimeric genomic fragments, vector or the transposon concatamer and then presents the clipped sequence reads to a sequence mapping server which aligns them to an Ensembl genome. Insertion sites can then be navigated in Ensembl in the context of genomic features such as gene structures. iMapper also generates test-based format for nucleic acid or protein sequences (FASTA) and generic file format (GFF) files of the clipped sequence reads and provides a graphical overview of the mapped insertion sites against a karyotype. iMapper is designed for high-throughput applications and can efficiently process thousands of DNA sequence reads. AVAILABILITY: iMapper is web based and can be accessed at http://www.sanger.ac.uk/cgi-bin/teams/team113/imapper.cgi.     

BM-map: Bayesian mapping of multireads for next-generation sequencing data

Next-generation sequencing (NGS) technology generates millions of short reads, which provide valuable information for various aspects of cellular activities and biological functions. A key step in NGS applications (e.g., RNA-Seq) is to map short reads to correct genomic locations within the source genome. While most reads are mapped to a unique location, a significant proportion of reads align to multiple genomic locations with equal or similar numbers of mismatches; these are called multireads. The ambiguity in mapping the multireads may lead to bias in downstream analyses. Currently, most practitioners discard the multireads in their analysis, resulting in a loss of valuable information, especially for the genes with similar sequences. To refine the read mapping, we develop a Bayesian model that computes the posterior probability of mapping a multiread to each competing location. The probabilities are used for downstream analyses, such as the quantification of gene expression. We show through simulation studies and RNA-Seq analysis of real life data that the Bayesian method yields better mapping than the current leading methods. We provide a C++ program for downloading that is being packaged into a user-friendly software.     

Improving Illumina assemblies with Hi‐C and long reads: An example with the North African dromedary

Researchers have assembled thousands of eukaryotic genomes using Illumina reads, but traditional mate‐pair libraries cannot span all repetitive elements, resulting in highly fragmented assemblies. However, both chromosome conformation capture techniques, such as Hi‐C and Dovetail Genomics Chicago libraries and long‐read sequencing, such as Pacific Biosciences and Oxford Nanopore, help span and resolve repetitive regions and therefore improve genome assemblies. One important livestock species of arid regions that does not have a high‐quality contiguous reference genome is the dromedary (Camelus dromedarius). Draft genomes exist but are highly fragmented, and a high‐quality reference genome is needed to understand adaptation to desert environments and artificial selection during domestication. Dromedaries are among the last livestock species to have been domesticated, and together with wild and domestic Bactrian camels, they are the only representatives of the Camelini tribe, which highlights their evolutionary significance. Here we describe our efforts to improve the North African dromedary genome. We used Chicago and Hi‐C sequencing libraries from Dovetail Genomics to resolve the order of previously assembled contigs, producing almost chromosome‐level scaffolds. Remaining gaps were filled with Pacific Biosciences long reads, and then scaffolds were comparatively mapped to chromosomes. Long reads added 99.32 Mbp to the total length of the new assembly. Dovetail Chicago and Hi‐C libraries increased the longest scaffold over 12‐fold, from 9.71 Mbp to 124.99 Mbp and the scaffold N50 over 50‐fold, from 1.48 Mbp to 75.02 Mbp. We demonstrate that Illumina de novo assemblies can be substantially upgraded by combining chromosome conformation capture and long‐read sequencing.     

Combination of de novo assembly of massive sequencing reads with classical repeat prediction improves identification of repetitive sequences in Schistosoma mansoni

The genome of the parasitic platyhelminth Schistosoma mansoni is composed of approximately 40% of repetitive sequences of which roughly 20% correspond to transposable elements. When the genome sequence became available, conventional repeat prediction programs were used to find these repeats, but only a fraction could be identified. To exhaustively characterize the repeats we applied a new massive sequencing based strategy: we re-sequenced the genome by next generation sequencing, aligned the sequencing reads to the genome and assembled all multiple-hit reads into contigs corresponding to the repetitive part of the genome. We present here, for the first time, this de novo repeat assembly strategy and we confirm that such assembly is feasible. We identified and annotated 4,143 new repeats in the S. mansoni genome. At least one third of the repeats are transcribed. This strategy allowed us also to identify 14 new microsatellite markers, which can be used for pedigree studies. Annotations and the combined (previously known and new) 5,420 repeat sequences (corresponding to 47% of the genome) are available for download (http://methdb.univ-perp.fr/downloads/).     

SMRT sequencing only de novo assembly of the sugar beet (Beta vulgaris) chloroplast genome

Background

Third generation sequencing methods, like SMRT (Single Molecule, Real-Time) sequencing developed by Pacific Biosciences, offer much longer read length in comparison to Next Generation Sequencing (NGS) methods. Hence, they are well suited for de novo- or re-sequencing projects. Sequences generated for these purposes will not only contain reads originating from the nuclear genome, but also a significant amount of reads originating from the organelles of the target organism. These reads are usually discarded but they can also be used for an assembly of organellar replicons. The long read length supports resolution of repetitive regions and repeats within the organelles genome which might be problematic when just using short read data. Additionally, SMRT sequencing is less influenced by GC rich areas and by long stretches of the same base.

Results

We describe a workflow for a de novo assembly of the sugar beet (Beta vulgaris ssp. vulgaris) chloroplast genome sequence only based on data originating from a SMRT sequencing dataset targeted on its nuclear genome. We show that the data obtained from such an experiment are sufficient to create a high quality assembly with a higher reliability than assemblies derived from e.g. Illumina reads only. The chloroplast genome is especially challenging for de novo assembling as it contains two large inverted repeat (IR) regions. We also describe some limitations that still apply even though long reads are used for the assembly.

Conclusions

SMRT sequencing reads extracted from a dataset created for nuclear genome (re)sequencing can be used to obtain a high quality de novo assembly of the chloroplast of the sequenced organism. Even with a relatively small overall coverage for the nuclear genome it is possible to collect more than enough reads to generate a high quality assembly that outperforms short read based assemblies. However, even with long reads it is not always possible to clarify the order of elements of a chloroplast genome sequence reliantly which we could demonstrate with Fosmid End Sequences (FES) generated with Sanger technology. Nevertheless, this limitation also applies to short read sequencing data but is reached in this case at a much earlier stage during finishing.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-015-0726-6) contains supplementary material, which is available to authorized users.     

UnSplicer: mapping spliced RNA-seq reads in compact genomes and filtering noisy splicing

Accurate mapping of spliced RNA-Seq reads to genomic DNA has been known as a challenging problem. Despite significant efforts invested in developing efficient algorithms, with the human genome as a primary focus, the best solution is still not known. A recently introduced tool, TrueSight, has demonstrated better performance compared with earlier developed algorithms such as TopHat and MapSplice. To improve detection of splice junctions, TrueSight uses information on statistical patterns of nucleotide ordering in intronic and exonic DNA. This line of research led to yet another new algorithm, UnSplicer, designed for eukaryotic species with compact genomes where functional alternative splicing is likely to be dominated by splicing noise. Genome-specific parameters of the new algorithm are generated by GeneMark-ES, an ab initio gene prediction algorithm based on unsupervised training. UnSplicer shares several components with TrueSight; the difference lies in the training strategy and the classification algorithm. We tested UnSplicer on RNA-Seq data sets of Arabidopsis thaliana, Caenorhabditis elegans, Cryptococcus neoformans and Drosophila melanogaster. We have shown that splice junctions inferred by UnSplicer are in better agreement with knowledge accumulated on these well-studied genomes than predictions made by earlier developed tools.     

Investigating repetitively matching short sequencing reads: The enigmatic nature of H3K9me3

Most histone modifications can easily be characterized as either activating or repressive. For example, histone3, lysine 4 trimethylation (H3K4me3) is generally considered a distinct sign of actively transcribed promoters while H3K27me3 is generally found at repressed genes. This is not the case for H3K9me3, the subject of this communication, which is a modification that has traditionally been considered a mark of constitutive heterochromatin, but has also been found in significant levels in expressed genes. We therefore sought to use new high-throughput genome-wide maps of H3K9me3 localization to investigate the conflicting hypotheses concerning the nature of this modification. Before we could accurately analyze the locations of H3K9me3 along the genome, and especially in repetitive locations, we developed a method for accurately utilizing short sequencing reads that do not map uniquely to a location in the genome. Investigating the locations of H3K9me3 along the genome allowed us to determine that, while there are high levels of H3K9me3 outside of genes, this modification is not absent from genes. Therefore, we suggest that H3K9me3 may have a role in chromatin organization rather than being directly related to gene expression. In addition, we have found that there is a need to include repetitively matching reads in any high-throughput sequencing experiment.     

Genome-wide SNP discovery in walnut with an AGSNP pipeline updated for SNP discovery in allogamous organisms

ABSTRACT: BACKGROUND: A genome-wide set of single nucleotide polymorphisms (SNPs) is a valuable resource in genetic research and breeding and is usually developed by re-sequencing a genome. If a genome sequence is not available, an alternative strategy must be used. We previously reported the development of a pipeline (AGSNP) for genome-wide SNP discovery in coding sequences and other single-copy DNA without a complete genome sequence in self-pollinating (autogamous) plants. Here we updated this pipeline for SNP discovery in outcrossing (allogamous) species and demonstrated its efficacy in SNP discovery in walnut (Juglans regia L.). RESULTS: The first step in the original implementation of the AGSNP pipeline was the construction of a reference sequence and the identification of single-copy sequences in it. To identify single-copy sequences, multiple genome equivalents of short SOLiD reads of another individual were mapped to shallow genome coverage of long Sanger or Roche 454 reads making up the reference sequence. The relative depth of SOLiD reads was used to filter out repeated sequences from single-copy sequences in the reference sequence. The second step was a search for SNPs between SOLiD reads and the reference sequence. Polymorphism within the mapped SOLiD reads would have precluded SNP discovery; hence both individuals had to be homozygous. The AGSNP pipeline was updated here for using SOLiD or other type of short reads of a heterozygous individual for these two principal steps. A total of 32.6X walnut genome equivalents of SOLiD reads of vegetatively propagated walnut scion cultivar 'Chandler' were mapped to 48,661 'Chandler' bacterial artificial chromosome (BAC) end sequences (BESs) produced by Sanger sequencing during the construction of a walnut physical map. A total of 22,799 putative SNPs were initially identified. A total of 6,000 Infinium II type SNPs evenly distributed along the walnut physical map were selected for the construction of an Infinium BeadChip, which was used to genotype a walnut mapping population having 'Chandler' as one of the parents. Genotyping results were used to adjust the filtering parameters of the updated AGSNP pipeline. With the adjusted filtering criteria, 69.6% of SNPs discovered with the updated pipeline were real and could be mapped on the walnut genetic map. A total of 13,439 SNPs were discovered by BES re-sequencing. BESs harboring SNPs were in 677 FPC contigs covering 98% of the physical map of the walnut genome. CONCLUSION: The updated AGSNP pipeline is a versatile SNP discovery tool for a high-throughput, genome-wide SNP discovery in both autogamous and allogamous species. With this pipeline, a large set of SNPs were identified in a single walnut cultivar.