Ultra-deep mutant spectrum profiling: improving sequencing accuracy using overlapping read pairs

Sequencing by binding rivals SMOR error-corrected sequencing by synthesis technology for accurate detection and quantification of minor (< 0.1%) subpopulation variants

BACKGROUND

Detecting very minor (< 1%) subpopulations using next-generation sequencing is a critical need for multiple applications, including the detection of drug resistant pathogens and somatic variant detection in oncology. A recently available sequencing approach termed 'sequencing by binding (SBB)' claims to have higher base calling accuracy data "out of the box." This paper evaluates the utility of using SBB for the detection of ultra-rare drug resistant subpopulations in Mycobacterium tuberculosis (Mtb) using a targeted amplicon assay and compares the performance of SBB to single molecule overlapping reads (SMOR) error corrected sequencing by synthesis (SBS) data.

RESULTS

SBS displayed an elevated error rate when compared to SMOR error-corrected SBS and SBB techniques. SMOR error-corrected SBS and SBB technologies performed similarly within the linear range studies and error rate studies.

CONCLUSIONS

With lower sequencing error rates within SBB sequencing, this technique looks promising for both targeted and unbiased whole genome sequencing, leading to the identification of minor (< 1%) subpopulations without the need for error correction methods.

Multiple Mutations Associated with Emergent Variants Can Be Detected as Low-Frequency Mutations in Early SARS-CoV-2 Pandemic Clinical Samples

Genetic analysis of intra-host viral populations provides unique insight into pre-emergent mutations that may contribute to the genotype of future variants. Clinical samples positive for SARS-CoV-2 collected in California during the first months of the pandemic were sequenced to define the dynamics of mutation emergence as the virus became established in the state. Deep sequencing of 90 nasopharyngeal samples showed that many mutations associated with the establishment of SARS-CoV-2 globally were present at varying frequencies in a majority of the samples, even those collected as the virus was first detected in the US. A subset of mutations that emerged months later in consensus sequences were detected as subconsensus members of intra-host populations. Spike mutations P681H, H655Y, and V1104L were detected prior to emergence in variant genotypes, mutations were detected at multiple positions within the furin cleavage site, and pre-emergent mutations were identified in the nucleocapsid and the envelope genes. Because many of the samples had a very high depth of coverage, a bioinformatics pipeline, "Mappgene", was established that uses both iVar and LoFreq variant calling to enable identification of very low-frequency variants. This enabled detection of a spike protein deletion present in many samples at low frequency and associated with a variant of concern.

Evaluating whole-genome sequencing quality metrics for enteric pathogen outbreaks

Background

Whole genome sequencing (WGS) has gained increasing importance in responses to enteric bacterial outbreaks. Common analysis procedures for WGS, single nucleotide polymorphisms (SNPs) and genome assembly, are highly dependent upon WGS data quality.

Methods

Raw, unprocessed WGS reads from Escherichia coli, Salmonella enterica, and Shigella sonnei outbreak clusters were characterized for four quality metrics: PHRED score, read length, library insert size, and ambiguous nucleotide composition. PHRED scores were strongly correlated with improved SNPs analysis results in E. coli and S. enterica clusters.

Results

Assembly quality showed only moderate correlations with PHRED scores and library insert size, and then only for Salmonella. To improve SNP analyses and assemblies, we compared seven read-healing pipelines to improve these four quality metrics and to see how well they improved SNP analysis and genome assembly. The most effective read healing pipelines for SNPs analysis incorporated quality-based trimming, fixed-width trimming, or both. The Lyve-SET SNPs pipeline showed a more marked improvement than the CFSAN SNP Pipeline, but the latter performed better on raw, unhealed reads. For genome assembly, SPAdes enabled significant improvements in healed E. coli reads only, while Skesa yielded no significant improvements on healed reads.

Conclusions

PHRED scores will continue to be a crucial quality metric albeit not of equal impact across all types of analyses for all enteric bacteria. While trimming-based read healing performed well for SNPs analyses, different read healing approaches are likely needed for genome assembly or other, emerging WGS analysis methodologies.

Drivers of within-host genetic diversity in acute infections of viruses

Genetic diversity is the fuel of evolution and facilitates adaptation to novel environments. However, our understanding of what drives differences in the genetic diversity during the early stages of viral infection is somewhat limited. Here, we use ultra-deep sequencing to interrogate 43 clinical samples taken from early infections of the human-infecting viruses HIV, RSV and CMV. Hundreds to thousands of virus templates were sequenced per sample, allowing us to reveal dramatic differences in within-host genetic diversity among virus populations. We found that increased diversity was mostly driven by presence of multiple divergent genotypes in HIV and CMV samples, which we suggest reflect multiple transmitted/founder viruses. Conversely, we detected an abundance of low frequency hyper-edited genomes in RSV samples, presumably reflecting defective virus genomes (DVGs). We suggest that RSV is characterized by higher levels of cellular co-infection, which allow for complementation and hence elevated levels of DVGs.

The few days or weeks following infection with a virus, termed acute infection, are critical for virus establishment. Here we sought to characterize what leads to differences in the genetic diversity of different viruses sampled during acute infection. We performed ultra-deep sequencing of hundreds to thousands viral genomes from forty-three samples spanning three pathogenic human viruses: HIV, RSV and CMV. We found major differences in the genetic diversity of these different viruses, and in different patients infected with the same virus. We investigated the factors responsible for these differences. We found that the DNA virus CMV was less diverse, most likely since it has a lower mutation rate than the RNA viruses HIV and RSV. We also found that the samples with the highest genetic diversity, which included one CMV sample and two HIV samples, bore evidence for multiple genotype infection. In other words, patients from whom these samples were taken were infected with two different “strains” of the virus. Finally, we also found evidence that viral genomes of HIV, and in particular RSV, are edited by the innate immune system of the host, leading to the presence of defective virus genomes.

Applying next-generation sequencing to unravel the mutational landscape in viral quasispecies

Next-generation sequencing (NGS) has revolutionized the scale and depth of biomedical sciences. Because of its unique ability for the detection of sub-clonal variants within genetically diverse populations, NGS has been successfully applied to analyze and quantify the exceptionally-high diversity within viral quasispecies, and many low-frequency drug- or vaccine-resistant mutations of therapeutic importance have been discovered. Although many works have intensively discussed the latest NGS approaches and applications in general, none of them has focused on applying NGS in viral quasispecies studies, mostly due to the limited ability of current NGS technologies to accurately detect and quantify rare viral variants. Here, we summarize several error-correction strategies that have been developed to enhance the detection accuracy of minority variants. We also discuss critical considerations for preparing a sequencing library from viral RNAs and for analyzing NGS data to unravel the mutational landscape.

A Bayesian model integration for mutation calling through data partitioning

Motivation

Detection of somatic mutations from tumor and matched normal sequencing data has become among the most important analysis methods in cancer research. Some existing mutation callers have focused on additional information, e.g. heterozygous single-nucleotide polymorphisms (SNPs) nearby mutation candidates or overlapping paired-end read information. However, existing methods cannot take multiple information sources into account simultaneously. Existing Bayesian hierarchical model-based methods construct two generative models, the tumor model and error model, and limited information sources have been modeled.

Results

We proposed a Bayesian model integration framework named as partitioning-based model integration. In this framework, through introducing partitions for paired-end reads based on given information sources, we integrate existing generative models and utilize multiple information sources. Based on that, we constructed a novel Bayesian hierarchical model-based method named as OHVarfinDer. In both the tumor model and error model, we introduced partitions for a set of paired-end reads that cover a mutation candidate position, and applied a different generative model for each category of paired-end reads. We demonstrated that our method can utilize both heterozygous SNP information and overlapping paired-end read information effectively in simulation datasets and real datasets.

Availability and implementation

https://github.com/takumorizo/OHVarfinDer.

Supplementary information

Supplementary data are available at Bioinformatics online.

Measurement error and variant-calling in deep Illumina sequencing of HIV

MOTIVATION

Next-generation deep sequencing of viral genomes, particularly on the Illumina platform, is increasingly applied in HIV research. Yet, there is no standard protocol or method used by the research community to account for measurement errors that arise during sample preparation and sequencing. Correctly calling high and low-frequency variants while controlling for erroneous variants is an important precursor to downstream interpretation, such as studying the emergence of HIV drug-resistance mutations, which in turn has clinical applications and can improve patient care.

RESULTS

We developed a new variant-calling pipeline, hivmmer, for Illumina sequences from HIV viral genomes. First, we validated hivmmer by comparing it to other variant-calling pipelines on real HIV plasmid datasets. We found that hivmmer achieves a lower rate of erroneous variants, and that all methods agree on the frequency of correctly called variants. Next, we compared the methods on an HIV plasmid dataset that was sequenced using Primer ID, an amplicon-tagging protocol, which is designed to reduce errors and amplification bias during library preparation. We show that the Primer ID consensus exhibits fewer erroneous variants compared to the variant-calling pipelines, and that hivmmer more closely approaches this low error rate compared to the other pipelines. The frequency estimates from the Primer ID consensus do not differ significantly from those of the variant-calling pipelines.

AVAILABILITY AND IMPLEMENTATION

hivmmer is freely available for non-commercial use from https://github.com/kantorlab/hivmmer.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Multiscale analysis for patterns of Zika virus genotype emergence, spread, and consequence

The question of how Zika virus (ZIKV) changed from a seemingly mild virus to a human pathogen capable of microcephaly and sexual transmission remains unanswered. The unexpected emergence of ZIKV's pathogenicity and capacity for sexual transmission may be due to genetic changes, and future changes in phenotype may continue to occur as the virus expands its geographic range. Alternatively, the sheer size of the 2015-16 epidemic may have brought attention to a pre-existing virulent ZIKV phenotype in a highly susceptible population. Thus, it is important to identify patterns of genetic change that may yield a better understanding of ZIKV emergence and evolution. However, because ZIKV has an RNA genome and a polymerase incapable of proofreading, it undergoes rapid mutation which makes it difficult to identify combinations of mutations associated with viral emergence. As next generation sequencing technology has allowed whole genome consensus and variant sequence data to be generated for numerous virus samples, the task of analyzing these genomes for patterns of mutation has become more complex. However, understanding which combinations of mutations spread widely and become established in new geographic regions versus those that disappear relatively quickly is essential for defining the trajectory of an ongoing epidemic. In this study, multiscale analysis of the wealth of genomic data generated over the course of the epidemic combined with in vivo laboratory data allowed trends in mutations and outbreak trajectory to be assessed. Mutations were detected throughout the genome via deep sequencing, and many variants appeared in multiple samples and in some cases become consensus. Similarly, amino acids that were previously consensus in pre-outbreak samples were detected as low frequency variants in epidemic strains. Protein structural models indicate that most of the mutations associated with the epidemic transmission occur on the exposed surface of viral proteins. At the macroscale level, consensus data was organized into large and interactive databases to allow the spread of individual mutations and combinations of mutations to be visualized and assessed for temporal and geographical patterns. Thus, the use of multiscale modeling for identifying mutations or combinations of mutations that impact epidemic transmission and phenotypic impact can aid the formation of hypotheses which can then be tested using reverse genetics.

A likelihood ratio-based approach for improved source attribution in microbiological forensic investigations

A common objective in microbial forensic investigations is to identify the origin of a recovered pathogenic bacterium by DNA sequencing. However, there is currently no consensus about how degrees of belief in such origin hypotheses should be quantified, interpreted, and communicated to wider audiences. To fill this gap, we have developed a concept based on calculating probabilistic evidential values for microbial forensic hypotheses. The likelihood-ratio method underpinning this concept is widely used in other forensic fields, such as human DNA matching, where results are readily interpretable and have been successfully communicated in juridical hearings. The concept was applied to two case scenarios of interest in microbial forensics: (1) identifying source cultures among series of very similar cultures generated by parallel serial passage of the Tier 1 pathogen Francisella tularensis, and (2) finding the production facilities of strains isolated in a real disease outbreak caused by the human pathogen Listeria monocytogenes. Evidence values for the studied hypotheses were computed based on signatures derived from whole genome sequencing data, including deep-sequenced low-frequency variants and structural variants such as duplications and deletions acquired during serial passages. In the F. tularensis case study, we were able to correctly assign fictive evidence samples to the correct culture batches of origin on the basis of structural variant data. By setting up relevant hypotheses and using data on cultivated batch sources to define the reference populations under each hypothesis, evidential values could be calculated. The results show that extremely similar strains can be separated on the basis of amplified mutational patterns identified by high-throughput sequencing. In the L. monocytogenes scenario, analyses of whole genome sequence data conclusively assigned the clinical samples to specific sources of origin, and conclusions were formulated to facilitate communication of the findings. Taken together, these findings demonstrate the potential of using bacterial whole genome sequencing data, including data on both low frequency SNP signatures and structural variants, to calculate evidence values that facilitate interpretation and communication of the results. The concept could be applied in diverse scenarios, including both epidemiological and forensic source tracking of bacterial infectious disease outbreaks.

Subclonal mutation selection in mouse lymphomagenesis identifies known cancer loci and suggests novel candidates

Determining whether recurrent but rare cancer mutations are bona fide driver mutations remains a bottleneck in cancer research. Here we present the most comprehensive analysis of murine leukemia virus-driven lymphomagenesis produced to date, sequencing 700,000 mutations from >500 malignancies collected at time points throughout tumor development. This scale of data allows novel statistical approaches for identifying selected mutations and yields a high-resolution, genome-wide map of the selective forces surrounding cancer gene loci. We also demonstrate negative selection of mutations that may be deleterious to tumor development indicating novel avenues for therapy. Screening of two BCL2 transgenic models confirmed known drivers of human non-Hodgkin lymphoma, and implicates novel candidates including modifiers of immunosurveillance and MHC loci. Correlating mutations with genotypic and phenotypic features independently of local variance in mutation density also provides support for weakly evidenced cancer genes. An online resource http://mulvdb.org allows customized queries of the entire dataset.

Evidence implicating cancer drivers can be sparse when limited to clonal events. Here, the authors present a retrovirus driven in vivo lymphomagenesis time course including hundreds of thousands of subclonal mutations and demonstrate the utility of these in mapping the selective forces affecting cancer gene loci, including negatively selected mutations.

On statistical modeling of sequencing noise in high depth data to assess tumor evolution

One cause of cancer mortality is tumor evolution to therapy-resistant disease. First line therapy often targets the dominant clone, and drug resistance can emerge from preexisting clones that gain fitness through therapy-induced natural selection. Such mutations may be identified using targeted sequencing assays by analysis of noise in high-depth data. Here, we develop a comprehensive, unbiased model for sequencing error background. We find that noise in sufficiently deep DNA sequencing data can be approximated by aggregating negative binomial distributions. Mutations with frequencies above noise may have prognostic value. We evaluate our model with simulated exponentially expanded populations as well as data from cell line and patient sample dilution experiments, demonstrating its utility in prognosticating tumor progression. Our results may have the potential to identify significant mutations that can cause recurrence. These results are relevant in the pre-treatment clinical setting to determine appropriate therapy and prepare for potential recurrence pretreatment.

MERIT reveals the impact of genomic context on sequencing error rate in ultra-deep applications

BACKGROUND

Rapid progress in high-throughput sequencing (HTS) and the development of novel library preparation methods have improved the sensitivity of detecting mutations in heterogeneous samples, specifically in high-depth (> 500×) clinical applications. However, HTS methods are bounded by their technical and theoretical limitations and sequencing errors cannot be completely eliminated. Comprehensive quantification of the background noise can highlight both the efficiency and the limitations of any HTS methodology, and help differentiate true mutations at low abundance from artifacts.

RESULTS

We introduce MERIT (Mutation Error Rate Inference Toolkit), designed for in-depth quantification of erroneous substitutions and small insertions and deletions. MERIT incorporates an all-inclusive variant caller and considers genomic context, including the nucleotides immediately at 5 'and 3 ', thereby establishing error rates for 96 possible substitutions as well as four single-base and 16 double-base indels. We applied MERIT to ultra-deep sequencing data (1,300,000 ×) obtained from the amplification of multiple clinically relevant loci, and showed a significant relationship between error rates and genomic contexts. In addition to observing significant difference between transversion and transition rates, we identified variations of more than 100-fold within each error type at high sequencing depths. For instance, T >G transversions in trinucleotide GTCs occurred 133.5 ± 65.9 more often than those in ATAs. Similarly, C >T transitions in GCGs were observed at 73.8 ± 10.5 higher rate than those in TCTs. We also devised an in silico approach to determine the optimal sequencing depth, where errors occur at rates similar to those of expected true mutations. Our analyses showed that increasing sequencing depth might improve sensitivity for detecting some mutations based on their genomic context. For example, T >G rate of error in GTCs did not change when sequenced beyond 10,000 ×; in contrast, T >G rate in TTAs consistently improved even at above 500,000 ×.

CONCLUSIONS

Our results demonstrate significant variation in nucleotide misincorporation rates, and suggest that genomic context should be considered for comprehensive profiling of specimen-specific and sequencing artifacts in high-depth assays. This data provide strong evidence against assigning a single allele frequency threshold to call mutations, for it can result in substantial false positive as well as false negative variants, with important clinical consequences.

Ultrasensitive and high-efficiency screen of de novo low-frequency mutations by o2n-seq

Detection of de novo, low-frequency mutations is essential for characterizing cancer genomes and heterogeneous cell populations. However, the screening capacity of current ultrasensitive NGS methods is inadequate owing to either low-efficiency read utilization or severe amplification bias. Here, we present o2n-seq, an ultrasensitive and high-efficiency NGS library preparation method for discovering de novo, low-frequency mutations. O2n-seq reduces the error rate of NGS to 10-5-10-8. The efficiency of its data usage is about 10-30 times higher than that of barcode-based strategies. For detecting mutations with allele frequency (AF) 1% in 4.6 Mb-sized genome, the sensitivity and specificity of o2n-seq reach to 99% and 98.64%, respectively. For mutations with AF around 0.07% in phix174, o2n-seq detects all the mutations with 100% specificity. Moreover, we successfully apply o2n-seq to screen de novo, low-frequency mutations in human tumours. O2n-seq will aid to characterize the landscape of somatic mutations in research and clinical settings.

OVarCall: Bayesian Mutation Calling Method Utilizing Overlapping Paired-End Reads

Detection of somatic mutations from tumor and matched normal sequencing data has become a standard approach in cancer research. Although a number of mutation callers have been developed, it is still difficult to detect mutations with low allele frequency even in exome sequencing. We expect that overlapping paired-end read information is effective for this purpose, but no mutation caller has modeled overlapping information statistically in a proper form in exome sequence data. Here, we develop a Bayesian hierarchical method, OVar- Call (https://github.com/takumorizo/OVarCall), where overlapping paired-end read information improves the accuracy of low allele frequency mutation detection. Firstly, we construct two generative models: one is for reads with somatic variants generated from tumor cells and the other is for reads that does not have somatic variants but potentially includes sequence errors. Secondly, we calculate marginal likelihood for each model using a variational Bayesian algorithm to compute Bayes factor for the detection of somatic mutations. We empirically evaluated the performance of OVarCall and confirmed its better performance than other existing methods.

Extremely Rare Polymorphisms in Saccharomyces cerevisiae Allow Inference of the Mutational Spectrum

The characterization of mutational spectra is usually carried out in one of three ways-by direct observation through mutation accumulation (MA) experiments, through parent-offspring sequencing, or by indirect inference from sequence data. Direct observations of spontaneous mutations with MA experiments are limited, given (i) the rarity of spontaneous mutations, (ii) applicability only to laboratory model species with short generation times, and (iii) the possibility that mutational spectra under lab conditions might be different from those observed in nature. Trio sequencing is an elegant solution, but it is not applicable in all organisms. Indirect inference, usually from divergence data, faces no such technical limitations, but rely upon critical assumptions regarding the strength of natural selection that are likely to be violated. Ideally, new mutational events would be directly observed before the biased filter of selection, and without the technical limitations common to lab experiments. One approach is to identify very young mutations from population sequencing data. Here we do so by leveraging two characteristics common to all new mutations-new mutations are necessarily rare in the population, and absent in the genomes of immediate relatives. From 132 clinical yeast strains, we were able to identify 1,425 putatively new mutations and show that they exhibit extremely low signatures of selection, as well as display a mutational spectrum that is similar to that identified by a large scale MA experiment. We verify that population sequencing data are a potential wealth of information for inferring mutational spectra, and should be considered for analysis where MA experiments are infeasible or especially tedious.

Excess of mutational jackpot events in expanding populations revealed by spatial Luria-Delbrück experiments

The genetic diversity of growing cellular populations, such as biofilms, solid tumours or developing embryos, is thought to be dominated by rare, exceptionally large mutant clones. Yet, the emergence of these mutational jackpot events is only understood in well-mixed populations, where they stem from mutations that arise during the first few cell divisions. To study jackpot events in spatially structured populations, we track mutant clones in microbial populations using fluorescence microscopy and population sequencing. High-frequency mutations are found to be massively enriched in microbial colonies compared with well-shaken liquid cultures, as a result of late-occurring mutations surfing at the edge of range expansions. Thus, jackpot events can be generated not only when mutations arise early but also when they occur at favourable locations, which exacerbates their role in adaptation and disease. In particular, because spatial competition with the wild type keeps most mutant clones in a quiescent state, strong selection pressures that kill the wild type promote drug resistance.

Large mutant clones arising from early mutations in growing cell populations facilitate short-term evolution in microbes and in tumours. Here the authors analyse spatially expanding colonies, and show that large mutant clones can also arise late when they surf at expanding frontiers.

Estimating Exceptionally Rare Germline and Somatic Mutation Frequencies via Next Generation Sequencing

We used targeted next generation deep-sequencing (Safe Sequencing System) to measure ultra-rare de novo mutation frequencies in the human male germline by attaching a unique identifier code to each target DNA molecule. Segments from three different human genes (FGFR3, MECP2 and PTPN11) were studied. Regardless of the gene segment, the particular testis donor or the 73 different testis pieces used, the frequencies for any one of the six different mutation types were consistent. Averaging over the C>T/G>A and G>T/C>A mutation types the background mutation frequency was 2.6x10^-5 per base pair, while for the four other mutation types the average background frequency was lower at 1.5x10^-6 per base pair. These rates far exceed the well documented human genome average frequency per base pair (~10⁻⁸) suggesting a non-biological explanation for our data. By computational modeling and a new experimental procedure to distinguish between pre-mutagenic lesion base mismatches and a fully mutated base pair in the original DNA molecule, we argue that most of the base-dependent variation in background frequency is due to a mixture of deamination and oxidation during the first two PCR cycles. Finally, we looked at a previously studied disease mutation in the PTPN11 gene and could easily distinguish true mutations from the SSS background. We also discuss the limits and possibilities of this and other methods to measure exceptionally rare mutation frequencies, and we present calculations for other scientists seeking to design their own such experiments.

High-specificity detection of rare alleles with Paired-End Low Error Sequencing (PELE-Seq)

BACKGROUND

Polymorphic loci exist throughout the genomes of a population and provide the raw genetic material needed for a species to adapt to changes in the environment. The minor allele frequencies of rare Single Nucleotide Polymorphisms (SNPs) within a population have been difficult to track with Next-Generation Sequencing (NGS), due to the high error rate of standard methods such as Illumina sequencing.

RESULTS

We have developed a wet-lab protocol and variant-calling method that identifies both sequencing and PCR errors, called Paired-End Low Error Sequencing (PELE-Seq). To test the specificity and sensitivity of the PELE-Seq method, we sequenced control E. coli DNA libraries containing known rare alleles present at frequencies ranging from 0.2-0.4 % of the total reads. PELE-Seq had higher specificity and sensitivity than standard libraries. We then used PELE-Seq to characterize rare alleles in a Caenorhabditis remanei nematode worm population before and after laboratory adaptation, and found that minor and rare alleles can undergo large changes in frequency during lab-adaptation.

CONCLUSION

We have developed a method of rare allele detection that mitigates both sequencing and PCR errors, called PELE-Seq. PELE-Seq was evaluated using control E. coli populations and was then used to compare a wild C. remanei population to a lab-adapted population. The PELE-Seq method is ideal for investigating the dynamics of rare alleles in a broad range of reduced-representation sequencing methods, including targeted amplicon sequencing, RAD-Seq, ddRAD, and GBS. PELE-Seq is also well-suited for whole genome sequencing of mitochondria and viruses, and for high-throughput rare mutation screens.

Using ultra-sensitive next generation sequencing to dissect DNA damage-induced mutagenesis

Next generation sequencing (NGS) technologies have dramatically improved studies in biology and biomedical science. However, no optimal NGS approach is available to conveniently analyze low frequency mutations caused by DNA damage treatments. Here, by developing an exquisite ultra-sensitive NGS (USNGS) platform “EasyMF” and incorporating it with a widely used supF shuttle vector-based mutagenesis system, we can conveniently dissect roles of lesion bypass polymerases in damage-induced mutagenesis. In this improved mutagenesis analysis pipeline, the initial steps are the same as in the supF mutation assay, involving damaging the pSP189 plasmid followed by its transfection into human 293T cells to allow replication to occur. Then “EasyMF” is employed to replace downstream MBM7070 bacterial transformation and other steps for analyzing damage-induced mutation frequencies and spectra. This pipeline was validated by using UV damaged plasmid after its replication in lesion bypass polymerase-deficient 293T cells. The increased throughput and reduced cost of this system will allow us to conveniently screen regulators of translesion DNA synthesis pathway and monitor environmental genotoxic substances, which can ultimately provide insight into the mechanisms of genome stability and mutagenesis.

Sendai virus intra-host population dynamics and host immunocompetence influence viral virulence during in vivo passage

In vivo serial passage of non-pathogenic viruses has been shown to lead to increased viral virulence, and although the precise mechanism(s) are not clear, it is known that both host and viral factors are associated with increased pathogenicity. Under- or overnutrition leads to a decreased or dysregulated immune response and can increase viral mutant spectrum diversity and virulence. The objective of this study was to identify the role of viral mutant spectra dynamics and host immunocompetence in the development of pathogenicity during in vivo passage. Because the nutritional status of the host has been shown to affect the development of viral virulence, the diet of animal model reflected two extremes of diets which exist in the global population, malnutrition and obesity. Sendai virus was serially passaged in groups of mice with differing nutritional status followed by transmission of the passaged virus to a second host species, guinea pigs. Viral population dynamics were characterized using deep sequence analysis and computational modeling. Histopathology, viral titer and cytokine assays were used to characterize viral virulence. Viral virulence increased with passage and the virulent phenotype persisted upon passage to a second host species. Additionally, nutritional status of mice during passage influenced the phenotype. Sequencing revealed the presence of several non-synonymous changes in the consensus sequence associated with passage, a majority of which occurred in the hemagglutinin-neuraminidase and polymerase genes, as well as the presence of persistent high frequency variants in the viral population. In particular, an N1124D change in the consensus sequences of the polymerase gene was detected by passage 10 in a majority of the animals. In vivo comparison of an 1124D plaque isolate to a clone with 1124N genotype indicated that 1124D was associated with increased virulence.

Ultra-precise detection of mutations by droplet-based amplification of circularized DNA

Background

NGS (next generation sequencing) has been widely used in studies of biological processes, ranging from microbial evolution to cancer genomics. However, the error rate of NGS (0.1 % ~ 1 %) is still remaining a great challenge for comprehensively investigating the low frequency variations, and the current solution methods have suffered severe amplification bias or low efficiency.

Results

We creatively developed Droplet-CirSeq for relatively efficient, low-bias and ultra-sensitive identification of variations by combining millions of picoliter uniform-sized droplets with Cir-seq. Droplet-CirSeq is entitled with an incredibly low error rate of 3 ~ 5 X 10^-6. To systematically evaluate the performances of amplification uniformity and capability of mutation identification for Droplet-CirSeq, we took the mixtures of two E. coli strains as specific instances to simulate the circumstances of mutations with different frequencies. Compared with Cir-seq, the coefficient of variance of read depth for Droplet-CirSeq was 10 times less (p = 2.6 X 10^-3), and the identified allele frequency presented more concentrated to the authentic frequency of mixtures (p = 4.8 X 10^-3), illustrating a significant improvement of amplification bias and accuracy in allele frequency determination. Additionally, Droplet-CirSeq detected 2.5 times genuine SNPs (p < 0.001), achieved a 2.8 times lower false positive rate (p < 0.05) and a 1.5 times lower false negative rate (p < 0.001), in the case of a 3 pg DNA input. Intriguingly, the false positive sites predominantly represented in two types of base substitutions (G- > A, C- > T). Our findings indicated that 30 pg DNA input accommodated in 5 ~ 10 million droplets resulted in maximal detection of authentic mutations compared to 3 pg (p = 1.2 X 10^-8) and 300 pg input (p = 2.2 X 10^-3).

Conclusions

We developed a method namely Droplet-CirSeq to significantly improve the amplification bias, which presents obvious superiority over the currently prevalent methods in exploitation of ultra-low frequency mutations. Droplet-CirSeq would be promisingly used in the identification of low frequency mutations initiated from extremely low input DNA, such as DNA of uncultured microorganisms, captured DNA of target region, circulation DNA of plasma et al, and its creative conception of rolling circle amplification in droplets would also be used in other low input DNA amplification fields.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-016-2480-1) contains supplementary material, which is available to authorized users.

Middle East Respiratory Syndrome Coronavirus Intra-Host Populations Are Characterized by Numerous High Frequency Variants

Middle East respiratory syndrome coronavirus (MERS-CoV) is an emerging human pathogen related to SARS virus. In vitro studies indicate this virus may have a broad host range suggesting an increased pandemic potential. Genetic and epidemiological evidence indicate camels serve as a reservoir for MERS virus but the mechanism of cross species transmission is unclear and many questions remain regarding the susceptibility of humans to infection. Deep sequencing data was obtained from the nasal samples of three camels that had been experimentally infected with a human MERS-CoV isolate. A majority of the genome was covered and average coverage was greater than 12,000x depth. Although only 5 mutations were detected in the consensus sequences, 473 intrahost single nucleotide variants were identified. Many of these variants were present at high frequencies and could potentially influence viral phenotype and the sensitivity of detection assays that target these regions for primer or probe binding.

Highly sensitive, non-invasive detection of colorectal cancer mutations using single molecule, third generation sequencing

Colorectal cancer (CRC) represents one of the most prevalent and lethal malignant neoplasms and every individual of age 50 and above should undergo regular CRC screening. Currently, the most effective preventive screening procedure to detect adenomatous polyps, the precursors to CRC, is colonoscopy. Since every colorectal cancer starts as a polyp, detecting all polyps and removing them is crucial. By exactly doing that, colonoscopy reduces CRC incidence by 80%, however it is an invasive procedure that might have unpleasant and, in rare occasions, dangerous side effects. Despite numerous efforts over the past two decades, a non-invasive screening method for the general population with detection rates for adenomas and CRC similar to that of colonoscopy has not yet been established. Recent advances in next generation sequencing technologies have yet to be successfully applied to this problem, because the detection of rare mutations has been hindered by the systematic biases due to sequencing context and the base calling quality of NGS. We present the first study that applies the high read accuracy and depth of single molecule, real time, circular consensus sequencing (SMRT-CCS) to the detection of mutations in stool DNA in order to provide a non-invasive, sensitive and accurate test for CRC. In stool DNA isolated from patients diagnosed with adenocarcinoma, we are able to detect mutations at frequencies below 0.5% with no false positives. This approach establishes a foundation for a non-invasive, highly sensitive assay to screen the population for CRC and the early stage adenomas that lead to CRC.

Genome plasticity of triple-reassortant H1N1 influenza A virus during infection of vaccinated pigs

To gain insight into the evolution of influenza A viruses (IAVs) during infection of vaccinated pigs, we experimentally infected a 3-week-old naive pig with a triple-reassortant H1N1 IAV and placed the seeder pig in direct contact with a group of age-matched vaccinated pigs (n = 10). We indexed the genetic diversity and evolution of the virus at an intra-host level by deep sequencing the entire genome directly from nasal swabs collected at two separate samplings during infection. We obtained 13 IAV metagenomes from 13 samples, which included the virus inoculum and two samples from each of the six pigs that tested positive for IAV during the study. The infection produced a population of heterogeneous alleles (sequence variants) that was dynamic over time. Overall, 794 polymorphisms were identified amongst all samples, which yielded 327 alleles, 214 of which were unique sequences. A total of 43 distinct haemagglutinin proteins were translated, two of which were observed in multiple pigs, whereas the neuraminidase (NA) was conserved and only one dominant NA was found throughout the study. The genetic diversity of IAVs changed dynamically within and between pigs. However, most of the substitutions observed in the internal gene segments were synonymous. Our results demonstrated remarkable IAV diversity, and the complex, rapid and dynamic evolution of IAV during infection of vaccinated pigs that can only be appreciated with repeated sampling of individual animals and deep sequence analysis.

Ultradeep Sequencing for Detection of Quasispecies Variants in the Major Hydrophilic Region of Hepatitis B Virus in Indonesian Patients

Quasispecies of hepatitis B virus (HBV) with variations in the major hydrophilic region (MHR) of the HBV surface antigen (HBsAg) can evolve during infection, allowing HBV to evade neutralizing antibodies. These escape variants may contribute to chronic infections. In this study, we looked for MHR variants in HBV quasispecies using ultradeep sequencing and evaluated the relationship between these variants and clinical manifestations in infected patients. We enrolled 30 Indonesian patients with hepatitis B infection (11 with chronic hepatitis and 19 with advanced liver disease). The most common subgenotype/subtype of HBV was B3/adw (97%). The HBsAg titer was lower in patients with advanced liver disease than that in patients with chronic hepatitis. The MHR variants were grouped based on the percentage of the viral population affected: major, ≥20% of the total population; intermediate, 5% to <20%; and minor, 1% to <5%. The rates of MHR variation that were present in the major and intermediate viral population were significantly greater in patients with advanced liver disease than those in chronic patients. The most frequent MHR variants related to immune evasion in the major and intermediate populations were P120Q/T, T123A, P127T, Q129H/R, M133L/T, and G145R. The major population of MHR variants causing impaired of HBsAg secretion (e.g., G119R, Q129R, T140I, and G145R) was detected only in advanced liver disease patients. This is the first study to use ultradeep sequencing for the detection of MHR variants of HBV quasispecies in Indonesian patients. We found that a greater number of MHR variations was related to disease severity and reduced likelihood of HBsAg titer.

Best practices for evaluating single nucleotide variant calling methods for microbial genomics

Innovations in sequencing technologies have allowed biologists to make incredible advances in understanding biological systems. As experience grows, researchers increasingly recognize that analyzing the wealth of data provided by these new sequencing platforms requires careful attention to detail for robust results. Thus far, much of the scientific Communit’s focus for use in bacterial genomics has been on evaluating genome assembly algorithms and rigorously validating assembly program performance. Missing, however, is a focus on critical evaluation of variant callers for these genomes. Variant calling is essential for comparative genomics as it yields insights into nucleotide-level organismal differences. Variant calling is a multistep process with a host of potential error sources that may lead to incorrect variant calls. Identifying and resolving these incorrect calls is critical for bacterial genomics to advance. The goal of this review is to provide guidance on validating algorithms and pipelines used in variant calling for bacterial genomics. First, we will provide an overview of the variant calling procedures and the potential sources of error associated with the methods. We will then identify appropriate datasets for use in evaluating algorithms and describe statistical methods for evaluating algorithm performance. As variant calling moves from basic research to the applied setting, standardized methods for performance evaluation and reporting are required; it is our hope that this review provides the groundwork for the development of these standards.

Detection of Low-Level Mixed-Population Drug Resistance in Mycobacterium tuberculosis Using High Fidelity Amplicon Sequencing

Undetected and untreated, low-levels of drug resistant (DR) subpopulations in clinical Mycobacterium tuberculosis (Mtb) infections may lead to development of DR-tuberculosis, potentially resulting in treatment failure. Current phenotypic DR susceptibility testing has a theoretical potential for 1% sensitivity, is not quantitative, and requires several weeks to complete. The use of "single molecule-overlapping reads" (SMOR) analysis with next generation DNA sequencing for determination of ultra-rare target alleles in complex mixtures provides increased sensitivity over standard DNA sequencing. Ligation free amplicon sequencing with SMOR analysis enables the detection of resistant allele subpopulations at ≥0.1% of the total Mtb population in near real-time analysis. We describe the method using standardized mixtures of DNA from resistant and susceptible Mtb isolates and the assay's performance for detecting ultra-rare DR subpopulations in DNA extracted directly from clinical sputum samples. SMOR analysis enables rapid near real-time detection and tracking of previously undetectable DR sub-populations in clinical samples allowing for the evaluation of the clinical relevance of low-level DR subpopulations. This will provide insights into interventions aimed at suppressing minor DR subpopulations before they become clinically significant.

Distinguishing low frequency mutations from RT-PCR and sequence errors in viral deep sequencing data

BACKGROUND

RNA viruses have high mutation rates and exist within their hosts as large, complex and heterogeneous populations, comprising a spectrum of related but non-identical genome sequences. Next generation sequencing is revolutionising the study of viral populations by enabling the ultra deep sequencing of their genomes, and the subsequent identification of the full spectrum of variants within the population. Identification of low frequency variants is important for our understanding of mutational dynamics, disease progression, immune pressure, and for the detection of drug resistant or pathogenic mutations. However, the current challenge is to accurately model the errors in the sequence data and distinguish real viral variants, particularly those that exist at low frequency, from errors introduced during sequencing and sample processing, which can both be substantial.

RESULTS

We have created a novel set of laboratory control samples that are derived from a plasmid containing a full-length viral genome with extremely limited diversity in the starting population. One sample was sequenced without PCR amplification whilst the other samples were subjected to increasing amounts of RT and PCR amplification prior to ultra-deep sequencing. This enabled the level of error introduced by the RT and PCR processes to be assessed and minimum frequency thresholds to be set for true viral variant identification. We developed a genome-scale computational model of the sample processing and NGS calling process to gain a detailed understanding of the errors at each step, which predicted that RT and PCR errors are more likely to occur at some genomic sites than others. The model can also be used to investigate whether the number of observed mutations at a given site of interest is greater than would be expected from processing errors alone in any NGS data set. After providing basic sample processing information and the site's coverage and quality scores, the model utilises the fitted RT-PCR error distributions to simulate the number of mutations that would be observed from processing errors alone.

CONCLUSIONS

These data sets and models provide an effective means of separating true viral mutations from those erroneously introduced during sample processing and sequencing.

Rapid identification and recovery of ENU-induced mutations with next-generation sequencing and Paired-End Low-Error analysis

BACKGROUND

Targeting Induced Local Lesions IN Genomes (TILLING) is a reverse genetics approach to directly identify point mutations in specific genes of interest in genomic DNA from a large chemically mutagenized population. Classical TILLING processes, based on enzymatic detection of mutations in heteroduplex PCR amplicons, are slow and labor intensive.

RESULTS

Here we describe a new TILLING strategy in zebrafish using direct next generation sequencing (NGS) of 250 bp amplicons followed by Paired-End Low-Error (PELE) sequence analysis. By pooling a genomic DNA library made from over 9,000 N-ethyl-N-nitrosourea (ENU) mutagenized F1 fish into 32 equal pools of 288 fish, each with a unique Illumina barcode, we reduce the complexity of the template to a level at which we can detect mutations that occur in a single heterozygous fish in the entire library. MiSeq sequencing generates 250 base-pair overlapping paired-end reads, and PELE analysis aligns the overlapping sequences to each other and filters out any imperfect matches, thereby eliminating variants introduced during the sequencing process. We find that this filtering step reduces the number of false positive calls 50-fold without loss of true variant calls. After PELE we were able to validate 61.5% of the mutant calls that occurred at a frequency between 1 mutant call:100 wildtype calls and 1 mutant call:1000 wildtype calls in a pool of 288 fish. We then use high-resolution melt analysis to identify the single heterozygous mutation carrier in the 288-fish pool in which the mutation was identified.

CONCLUSIONS

Using this NGS-TILLING protocol we validated 28 nonsense or splice site mutations in 20 genes, at a two-fold higher efficiency than using traditional Cel1 screening. We conclude that this approach significantly increases screening efficiency and accuracy at reduced cost and can be applied in a wide range of organisms.

Limits of neutral drift: lessons from the in vitro evolution of two ribozymes

The relative contributions of adaptive selection and neutral drift to genetic change are unknown but likely depend on the inherent abundance of functional genotypes in sequence space and how accessible those genotypes are to one another. To better understand the relative roles of selection and drift in evolution, local fitness landscapes for two different RNA ligase ribozymes were examined using a continuous in vitro evolution system under conditions that foster the capacity for neutral drift to mediate genetic change. The exploration of sequence space was accelerated by increasing the mutation rate using mutagenic nucleotide analogs. Drift was encouraged by carrying out evolution within millions of separate compartments to exploit the founder effect. Deep sequencing of individuals from the evolved populations revealed that the distribution of genotypes did not escape the starting local fitness peak, remaining clustered around the sequence used to initiate evolution. This is consistent with a fitness landscape where high-fitness genotypes are sparse and well isolated, and suggests, at least in this context, that neutral drift alone is not a primary driver of genetic change. Neutral drift does, however, provide a repository of genetic variation upon which adaptive selection can act.

Ultra-deep sequencing of intra-host rabies virus populations during cross-species transmission

One of the hurdles to understanding the role of viral quasispecies in RNA virus cross-species transmission (CST) events is the need to analyze a densely sampled outbreak using deep sequencing in order to measure the amount of mutation occurring on a small time scale. In 2009, the California Department of Public Health reported a dramatic increase (350) in the number of gray foxes infected with a rabies virus variant for which striped skunks serve as a reservoir host in Humboldt County. To better understand the evolution of rabies, deep-sequencing was applied to 40 unpassaged rabies virus samples from the Humboldt outbreak. For each sample, approximately 11 kb of the 12 kb genome was amplified and sequenced using the Illumina platform. Average coverage was 17,448 and this allowed characterization of the rabies virus population present in each sample at unprecedented depths. Phylogenetic analysis of the consensus sequence data demonstrated that samples clustered according to date (1995 vs. 2009) and geographic location (northern vs. southern). A single amino acid change in the G protein distinguished a subset of northern foxes from a haplotype present in both foxes and skunks, suggesting this mutation may have played a role in the observed increased transmission among foxes in this region. Deep-sequencing data indicated that many genetic changes associated with the CST event occurred prior to 2009 since several nonsynonymous mutations that were present in the consensus sequences of skunk and fox rabies samples obtained from 20032010 were present at the sub-consensus level (as rare variants in the viral population) in skunk and fox samples from 1995. These results suggest that analysis of rare variants within a viral population may yield clues to ancestral genomes and identify rare variants that have the potential to be selected for if environment conditions change.

Understanding the role of genetic variants within a viral population is a necessary step toward predicting and treating emerging infectious diseases. The high mutation rate of RNA viruses increases the ability of these viruses to adapt to diverse hosts and cause new human and zoonotic diseases. The genetic diversity of a viral population within a host may allow the virus to adapt to a diverse array of selective pressures and enable cross-species transmission events. In 2009 a large outbreak of rabies in Northern California involved a skunk rabies virus variant that efficiently transmitted within a population of gray foxes, suggesting possible adaptation to a novel host species. To better understand the evolution of rabies virus that enabled this host jump, we applied deep-sequencing analysis to rabies virus samples from the outbreak. Deep-sequencing data indicated that many of the genetic changes associated with host jump occurred prior to 2009, and these mutations were present at very low frequencies in viral populations from samples dating back to 1995. These results suggest deep sequencing is useful for characterization of viral populations, and may provide insight to ancestral genomes and role of rare variants in viral emergence.

Direct assessment of transcription fidelity by high-resolution RNA sequencing

Cancerous and aging cells have long been thought to be impacted by transcription errors that cause genetic and epigenetic changes. Until now, a lack of methodology for directly assessing such errors hindered evaluation of their impact to the cells. We report a high-resolution Illumina RNA-seq method that can assess noncoded base substitutions in mRNA at 10⁻⁴–10⁻⁵ per base frequencies in vitro and in vivo. Statistically reliable detection of changes in transcription fidelity through ∼10³ nt DNA sites assures that the RNA-seq can analyze the fidelity in a large number of the sites where errors occur. A combination of the RNA-seq and biochemical analyses of the positions for the errors revealed two sequence-specific mechanisms that increase transcription fidelity by Escherichia coli RNA polymerase: (i) enhanced suppression of nucleotide misincorporation that improves selectivity for the cognate substrate, and (ii) increased backtracking of the RNA polymerase that decreases a chance of error propagation to the full-length transcript after misincorporation and provides an opportunity to proofread the error. This method is adoptable to a genome-wide assessment of transcription fidelity.