Alignment Of Sequencing Reads Research Articles

Unravelling reference bias in ancient DNA datasets.

The alignment of sequencing reads is a critical step in the characterization of ancient genomes. However, reference bias and spurious mappings pose a significant challenge, particularly as cutting-edge wet lab methods generate datasets that push the boundaries of alignment tools. Reference bias occurs when reference alleles are favoured over alternative alleles during mapping, whereas spurious mappings stem from either contamination or when endogenous reads fail to align to their correct position. Previous work has shown that these phenomena are correlated with read length but a more thorough investigation of reference bias and spurious mappings for ancient DNA has been lacking. Here, we use a range of empirical and simulated palaeogenomic datasets to investigate the impacts of mapping tools, quality thresholds, and reference genome on mismatch rates across read lengths. For these analyses, we introduce AMBER, a new bioinformatics tool for assessing the quality of ancient DNA mapping directly from BAM-files and informing on reference bias, read length cut-offs and reference selection. AMBER rapidly and simultaneously computes the sequence read mapping bias in the form of the mismatch rates per read length, cytosine deamination profiles at both CpG and non-CpG sites, fragment length distributions, and genomic breadth and depth of coverage. Using AMBER, we find that mapping algorithms and quality threshold choices dictate reference bias and rates of spurious alignment at different read lengths in a predictable manner, suggesting that optimized mapping parameters for each read length will be a key step in alleviating reference bias and spurious mappings. AMBER is available for noncommercial use on GitHub (https://github.com/tvandervalk/AMBER.git). Scripts used to generate and analyse simulated datasets are available on Github (https://github.com/sdolenz/refbias_scripts).

On the core segmentation algorithms of copy number variation detection tools.

Shotgun sequencing is a high-throughput method used to detect copy number variants (CNVs). Although there are numerous CNV detection tools based on shotgun sequencing, their quality varies significantly, leading to performance discrepancies. Therefore, we conducted a comprehensive analysis of next-generation sequencing-based CNV detection tools over the past decade. Our findings revealed that the majority of mainstream tools employ similar detection rationale: calculates the so-called read depth signal from aligned sequencing reads and then segments the signal by utilizing either circular binary segmentation (CBS) or hidden Markov model (HMM). Hence, we compared the performance of those two core segmentation algorithms in CNV detection, considering varying sequencing depths, segment lengths and complex types of CNVs. To ensure a fair comparison, we designed a parametrical model using mainstream statistical distributions, which allows for pre-excluding bias correction such as guanine-cytosine (GC) content during the preprocessing step. The results indicate the following key points: (1) Under ideal conditions, CBS demonstrates high precision, while HMM exhibits a high recall rate. (2) For practical conditions, HMM is advantageous at lower sequencing depths, while CBS is more competitive in detecting small variant segments compared to HMM. (3) In case involving complex CNVs resembling real sequencing, HMM demonstrates more robustness compared with CBS. (4) When facing large-scale sequencing data, HMM costs less time compared with the CBS, while their memory usage is approximately equal. This can provide an important guidance and reference for researchers to develop new tools for CNV detection.

Small polymorphisms are a source of ancestral bias in structural variant breakpoint placement.

High-quality genome assemblies and sophisticated algorithms have increased sensitivity for a wide range of variant types, and breakpoint accuracy for structural variants (SVs, ≥50 bp) has improved to near base pair precision. Despite these advances, many SV breakpoint locations are subject to systematic bias affecting variant representation. To understand why SV breakpoints are inconsistent across samples, we reanalyzed 64 phased haplotypes constructed from long-read assemblies released by the Human Genome Structural Variation Consortium (HGSVC). We identify 882 SV insertions and 180 SV deletions with variable breakpoints not anchored in tandem repeats (TRs) or segmental duplications (SDs). SVs called from aligned sequencing reads increase breakpoint disagreements by 2×-16×. Sequence accuracy had a minimal impact on breakpoints, but we observe a strong effect of ancestry. We confirm that SNP and indel polymorphisms are enriched at shifted breakpoints and are also absent from variant callsets. Breakpoint homology increases the likelihood of imprecise SV calls and the distance they are shifted, and tandem duplications are the most heavily affected SVs. Because graph genome methods normalize SV calls across samples, we investigated graphs generated by two different methods and find the resulting breakpoints are subject to other technical biases affecting breakpoint accuracy. The breakpoint inconsistencies we characterize affect ∼5% of the SVs called in a human genome and can impact variant interpretation and annotation. These limitations underscore a need for algorithm development to improve SV databases, mitigate the impact of ancestry on breakpoints, and increase the value of callsets for investigating breakpoint features.

Approaches for the Analysis and Interpretation of Whole-Genome Bisulfite Sequencing Data.

DNA methylation is a covalent modification of DNA that plays important roles in processes such as the regulation of gene expression, transcription factor binding, and suppression of transposable elements. The use of whole-genome bisulfite sequencing (WGBS) enables the genome-wide identification and quantification of DNA methylation patterns at single-base resolution and is the gold standard for the analysis of DNA methylation. However, the computational analysis of WGBS data can be particularly challenging, as many computationally intensive steps are required. Here, we outline step-by-step an approach for the analysis and interpretation of WGBS data. First, sequencing reads must be trimmed, quality-checked, and aligned to the genome. Second, DNA methylation levels are estimated at each cytosine position using the aligned sequence reads of the bisulfite-treated DNA. Third, regions of differential cytosine methylation between samples can be identified. Finally, these data need to be visualized and interpreted in the context of the biological question at hand.

Short Sequence Aligner Benchmarking for Chromatin Research.

Much of today's molecular science revolves around next-generation sequencing. Frequently, the first step in analyzing such data is aligning sequencing reads to a reference genome. This step is often taken for granted, but any analysis downstream of the alignment will be affected by the aligner's ability to correctly map sequences. In most cases, for research into chromatin structure and nucleosome positioning, ATAC-seq, ChIP-seq, and MNase-seq experiments use short read lengths. How well aligners manage these reads is critical. Most aligner programs will output mapped reads and unmapped reads. However, from a biological point of view, reads will fall into one of three categories: correctly mapped, incorrectly mapped, and unmapped. While increased sequencing depth can often compensate for unmapped reads, incorrectly and correctly mapped reads appear algorithmically identical but can produce biologically significant alterations in the results. For this reason, we are benchmarking various alignment programs to determine their propensity to incorrectly map short reads. As short-read alignment is an important step in ATAC-seq, ChIP-seq, and MNase-seq experiments, caution should be taken in mapping reads to ensure that the most accurate conclusions can be made from the data generated. Our analysis is intended to help investigators new to the field pick the alignment program best suited for their experimental conditions. In general, the aligners we tested performed well. BWA, Bowtie2, and Chromap were all exceptionally accurate, and we recommend using them. Furthermore, we show that longer read lengths do in fact lead to more accurate mappings.

Closing the Gap - Detection of 5q-Spinal Muscular Atrophy by Short-Read Next-Generation Sequencing and Unexpected Results in a Diagnostic Patient Cohort.

The importance of early diagnosis of 5q-Spinal muscular atrophy (5q-SMA) has heightened as early intervention can significantly improve clinical outcomes. In 96% of cases, 5q-SMA is caused by a homozygous deletion of SMN1. Around 4 % of patients carry a SMN1 deletion and a single-nucleotide variant (SNV) on the other allele. Traditionally, diagnosis is based on multiplex ligation probe amplification (MLPA) to detect homozygous or heterozygous exon 7 deletions in SMN1. Due to high homologies within the SMN1/SMN2 locus, sequence analysis to identify SNVs of the SMN1 gene is unreliable by standard Sanger or short-read next-generation sequencing (srNGS) methods. The objective was to overcome the limitations in high-throughput srNGS with the aim of providing SMA patients with a fast and reliable diagnosis to enable their timely therapy. A bioinformatics workflow to detect homozygous SMN1 deletions and SMN1 SNVs on srNGS analysis was applied to diagnostic whole exome and panel testing for suggested neuromuscular disorders (1684 patients) and to fetal samples in prenatal diagnostics (260 patients). SNVs were detected by aligning sequencing reads from SMN1 and SMN2 to an SMN1 reference sequence. Homozygous SMN1 deletions were identified by filtering sequence reads for the ,, gene-determining variant" (GDV). 10 patients were diagnosed with 5q-SMA based on (i) SMN1 deletion and hemizygous SNV (2 patients), (ii) homozygous SMN1 deletion (6 patients), and (iii) compound heterozygous SNVs in SMN1 (2 patients). Applying our workflow in srNGS-based panel and whole exome sequencing (WES) is crucial in a clinical laboratory, as otherwise patients with an atypical clinical presentation initially not suspected to suffer from SMA remain undiagnosed.

DeNovoCNN: a deep learning approach to de novo variant calling in next generation sequencing data.

De novo mutations (DNMs) are an important cause of genetic disorders. The accurate identification of DNMs from sequencing data is therefore fundamental to rare disease research and diagnostics. Unfortunately, identifying reliable DNMs remains a major challenge due to sequence errors, uneven coverage, and mapping artifacts. Here, we developed a deep convolutional neural network (CNN) DNM caller (DeNovoCNN), that encodes the alignment of sequence reads for a trio as 160n}{} times164 resolution images. DeNovoCNN was trained on DNMs of 5616 whole exome sequencing (WES) trios achieving total 96.74% recall and 96.55% precision on the test dataset. We find that DeNovoCNN has increased recall/sensitivity and precision compared to existing DNM calling approaches (GATK, DeNovoGear, DeepTrio, Samtools) based on the Genome in a Bottle reference dataset and independent WES and WGS trios. Validations of DNMs based on Sanger and PacBio HiFi sequencing confirm that DeNovoCNN outperforms existing methods. Most importantly, our results suggest that DeNovoCNN is likely robust against different exome sequencing and analyses approaches, thereby allowing the application on other datasets. DeNovoCNN is freely available as a Docker container and can be run on existing alignment (BAM/CRAM) and variant calling (VCF) files from WES and WGS without a need for variant recalling.

An SGSGeneloss-Based Method for Constructing a Gene Presence-Absence Table Using Mosdepth.

Presence-absence variants (PAV) are genomic regions present in some individuals of a species, but not others. PAVs have been shown to contribute to genomic diversity, especially in bacteria and plants. These structural variations have been linked to traits and can be used to track a species' evolutionary history. PAVs are usually called by aligning short read sequence data from one or more individuals to a reference genome or pangenome assembly, and then comparing coverage. Regions where reads do not align define absence in that individual, and the regions are classified as PAVs. The method below details how to align sequence reads to a reference and how to use the sequencing-coverage calculator Mosdepth to identify PAVs and construct a PAV table for use in downstream comparative genome analysis.

Abstract 707: Detection of SARS-CoV-2 co-infections with seasonal respiratory viruses using a multiplexed amplicon NGS panel

Abstract The ongoing COVID-19 pandemic, caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has affected everyone around the world. Most symptoms of COVID-19 overlap with symptoms of the flu and the common cold. With the approaching flu season, early and accurate detection of these viruses are critical not only to slow the spread of SARS-CoV-2, but also to ensure that patients receive the appropriate care based on their diagnosis, and to keep the healthcare system from becoming overloaded. We designed a next-generation sequencing (NGS) assay to detect the SARS-CoV-2, influenza A H1N1, influenza A H3N2, influenza B, RSV A, and RSV B viruses, and to allow for surveillance and subtype identification. The primers for this multiplex PCR-based method interrogate nearly every gene of each virus. Subsequent sequencing enables not only virus detection but also co-infection detection for these respiratory viruses. In addition, sequencing of virus-positive samples makes it possible to use this assay for mutation and strain analysis. Using the CleanPlex target enrichment technology, we prepared targeted amplicon libraries with human total RNA (50ng) spiked with SARS-CoV-2 synthetic RNA, influenza A synthetic RNA, influenza B synthetic RNA, RSV A RNA, and RSV B RNA (100-10,000 viral genome copies), either individually or in combination. The NGS library preparation protocol includes a reverse transcription step to generate cDNA, followed by a multiplexed PCR step for specific amplification of viral targets. A critical subsequent background cleaning step removes non-specific PCR products, and a final PCR adds sample indices and Illumina sequencing adapters to complete the NGS library preparation. Libraries were sequenced on the Illumina iSeq100 platform, and sequencing metrics as well as virus characterization were determined using Paragon Genomics's bioinformatics pipeline. Each sample was sequenced at 1,000 average paired-end reads per amplicon. After aligning sequencing reads to each viral genome, we were able to detect nearly every gene of each virus with high specificity. At viral loads of 10,000 copies, 96-100% of amplicons were covered at >100X, and mapped to the intended virus species as well as to the intended subtypes for influenza A and RSV. The assay also accurately identified co-infected samples, which contained equal copies of SARS-CoV-2 and influenza (either A H1N1, A H3N2, or B). In summary, the assay detects multi-target respiratory viruses with high accuracy and specificity. We plan to expand on our findings for this assay's performance with additional targets, co-infection studies and mutational analysis studies. Citation Format: Lucie S. Lee, Rounak Feigelman, David Debruyne, Julia Spencer, Chenyu Li, Zhitong Liu, Guoying Liu. Detection of SARS-CoV-2 co-infections with seasonal respiratory viruses using a multiplexed amplicon NGS panel [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 707.

Novel functional sequences uncovered through a bovine multiassembly graph

Many genomic analyses start by aligning sequencing reads to a linear reference genome. However, linear reference genomes are imperfect, lacking millions of bases of unknown relevance and are unable to reflect the genetic diversity of populations. This makes reference-guided methods susceptible to reference-allele bias. To overcome such limitations, we build a pangenome from six reference-quality assemblies from taurine and indicine cattle as well as yak. The pangenome contains an additional 70,329,827 bases compared to the Bos taurus reference genome. Our multiassembly approach reveals 30 and 10.1 million bases private to yak and indicine cattle, respectively, and between 3.3 and 4.4 million bases unique to each taurine assembly. Utilizing transcriptomes from 56 cattle, we show that these nonreference sequences encode transcripts that hitherto remained undetected from the B. taurus reference genome. We uncover genes, primarily encoding proteins contributing to immune response and pathogen-mediated immunomodulation, differentially expressed between Mycobacterium bovis-infected and noninfected cattle that are also undetectable in the B. taurus reference genome. Using whole-genome sequencing data of cattle from five breeds, we show that reads which were previously misaligned against the Bos taurus reference genome now align accurately to the pangenome sequences. This enables us to discover 83,250 polymorphic sites that segregate within and between breeds of cattle and capture genetic differentiation across breeds. Our work makes a so-far unused source of variation amenable to genetic investigations and provides methods and a framework for establishing and exploiting a more diverse reference genome.

Reference flow: reducing reference bias using multiple population genomes

Most sequencing data analyses start by aligning sequencing reads to a linear reference genome, but failure to account for genetic variation leads to reference bias and confounding of results downstream. Other approaches replace the linear reference with structures like graphs that can include genetic variation, incurring major computational overhead. We propose the reference flow alignment method that uses multiple population reference genomes to improve alignment accuracy and reduce reference bias. Compared to the graph aligner vg, reference flow achieves a similar level of accuracy and bias avoidance but with 14% of the memory footprint and 5.5 times the speed.

NGPINT: a next-generation protein-protein interaction software.

Mapping protein-protein interactions at a proteome scale is critical to understanding how cellular signaling networks respond to stimuli. Since eukaryotic genomes encode thousands of proteins, testing their interactions one-by-one is a challenging prospect. High-throughput yeast-two hybrid (Y2H) assays that employ next-generation sequencing to interrogate complementary DNA (cDNA) libraries represent an alternative approach that optimizes scale, cost and effort. We present NGPINT, a robust and scalable software to identify all putative interactors of a protein using Y2H in batch culture. NGPINT combines diverse tools to align sequence reads to target genomes, reconstruct prey fragments and compute gene enrichment under reporter selection. Central to this pipeline is the identification of fusion reads containing sequences derived from both the Y2H expression plasmid and the cDNA of interest. To reduce false positives, these fusion reads are evaluated as to whether the cDNA fragment forms an in-frame translational fusion with the Y2H transcription factor. NGPINT successfully recognized 95% of interactions in simulated test runs. As proof of concept, NGPINT was tested using published data sets and it recognized all validated interactions. NGPINT can process interaction data from any biosystem with an available genome or transcriptome reference, thus facilitating the discovery of protein-protein interactions in model and non-model organisms.

Abstract 4420: Polytect, a robust algorithm to detect somatic mutations in polymorphic MHC/HLA genes

Abstract Background: The immune system is able to recognize tumor cells through the presentation of mutant peptides via the major histocompatibility complex (MHC). Somatic mutations and loss of heterozygosity of MHC genes are a known mechanism of escape from this immune surveillance. Classical methods to identify somatic mutations involve alignment of sequencing reads to a standard reference genome, but this approach fails for the MHC genes. High sequence divergence between haplotypes and the GRCh38 reference, in addition to a large number of closely-related homologs, results in poor alignment and difficulty in variant detection. As a result, there is a lack of validated tools that can reliably identify mutations in the MHC genes, leaving this area understudied. Methods: We developed Polytect, a genomics tool that can call somatic mutations in highly polymorphic genes. Polytect accurately partitions reads across homologues, and aligns them against references that are dynamically constructed based on an individual's haplotype. This minimizes alignment errors, and enables the use of traditional variant callers. Reads are accurately partitioned among genes by exploring the space of alternative alignments to all haplotypes and pseudogenes. For the MHC, references are dynamically selected from the IMGT database of known haplotypes using a novel algorithm that models alignment to closely related sequences as a linear system. This allows us to solve for the true underlying haplotypes for each patient. Personalized references are then created by imputing germline mutations onto the selected haplotypes. Once reads are aligned to the personalized reference, somatic mutations can be called using existing methods. Results: In a set of validation experiments, selected references had a median levenshtein distance of 0 SNPs from the known ground truth sequences. However, suboptimal reference selection (> 15 SNPs) was more likely for genes with worse haplotype representation in IMGT and those with extremely variable intron lengths. Polytect achieved a sensitivity of 88%, detecting 65 out of 74 MHC-Class 1 mutation from a gold-standard dataset. We applied Polytect to a cohort of 706 patients with metastatic cancers from the MI-ONCOSEQ study. We found non-synonymous mutations in the MHC class 1 genes in 7% of tumors, a rate higher than the 3.3% reported in a previous study looking at primary tumors. We also applied Polytect to the previously uninvestigated MHC class 2 genes and found mutations in 6% of tumors. MHC mutations were most common in squamous cell carcinomas (40%, n = 20), melanomas (33%, n = 6), and colorectal cancers (25%, n = 8), compared to, for example, (10%, n=256) in prostate cancer. Conclusions: Polytect is an accurate and efficient algorithm to detect somatic mutations in polymorphic genes. Across a large cohort of metastatic tumors Polytect revealed a high prevalence (13%) of mutations in MHC class 1 and class 2 genes. Citation Format: Michael Brodie Mumphrey, Abhijit Parolia, Chandan Kumar-Sinha, Marcin Cieslik. Polytect, a robust algorithm to detect somatic mutations in polymorphic MHC/HLA genes [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4420.

Comparison of Read Mapping and Variant Calling Tools for the Analysis of Plant NGS Data.

High-throughput sequencing technologies have rapidly developed during the past years and have become an essential tool in plant sciences. However, the analysis of genomic data remains challenging and relies mostly on the performance of automatic pipelines. Frequently applied pipelines involve the alignment of sequence reads against a reference sequence and the identification of sequence variants. Since most benchmarking studies of bioinformatics tools for this purpose have been conducted on human datasets, there is a lack of benchmarking studies in plant sciences. In this study, we evaluated the performance of 50 different variant calling pipelines, including five read mappers and ten variant callers, on six real plant datasets of the model organism Arabidopsis thaliana. Sets of variants were evaluated based on various parameters including sensitivity and specificity. We found that all investigated tools are suitable for analysis of NGS data in plant research. When looking at different performance metrics, BWA-MEM and Novoalign were the best mappers and GATK returned the best results in the variant calling step.

SRPRISM (Single Read Paired Read Indel Substitution Minimizer): an efficient aligner for assemblies with explicit guarantees.

Alignment of sequence reads generated by next-generation sequencing is an integral part of most pipelines analyzing next-generation sequencing data. A number of tools designed to quickly align a large volume of sequences are already available. However, most existing tools lack explicit guarantees about their output. They also do not support searching genome assemblies, such as the human genome assembly GRCh38, that include primary and alternate sequences and placement information for alternate sequences to primary sequences in the assembly. This paper describes SRPRISM (Single Read Paired Read Indel Substitution Minimizer), an alignment tool for aligning reads without splices. SRPRISM has features not available in most tools, such as (i) support for searching genome assemblies with alternate sequences, (ii) partial alignment of reads with a specified region of reads to be included in the alignment, (iii) choice of ranking schemes for alignments, and (iv) explicit criteria for search sensitivity. We compare the performance of SRPRISM to GEM, Kart, STAR, BWA-MEM, Bowtie2, Hobbes, and Yara using benchmark sets for paired and single reads of lengths 100 and 250bp generated using DWGSIM. SRPRISM found the best results for most benchmark sets with error rate of up to ∼2.5% and GEM performed best for higher error rates. SRPRISM was also more sensitive than other tools even when sensitivity was reduced to improve run time performance. We present SRPRISM as a flexible read mapping tool that provides explicit guarantees on results.

What human sperm RNA-Seq tells us about the microbiome.

The study was designed to assess the capacity of human sperm RNA-seq data to gauge the diversity of the associated microbiome within the ejaculate. Semen samples were collected, and semen parameters evaluated at time of collection. Sperm RNA was isolated and subjected to RNA-seq. Microbial composition was determined by aligning sequencing reads not mapped to the human genome to the NCBI RefSeq bacterial, viral and archaeal genomes following RNA-Seq. Analysis of microbial assignments utilized phyloseq and vegan. Microbial composition within each sample was characterized as a function of microbial associated RNAs. Bacteria known to be associated with the male reproductive tract were present at similar levels in all samples representing 11 genera from four phyla with one exception, an outlier. Shannon diversity index (p < 0.001) and beta diversity (unweighted UniFrac distances, p = 9.99e-4; beta dispersion, p = 0.006) indicated the outlier was significantly different from all other samples. The outlier sample exhibited a dramatic increase in Streptococcus. Multiple testing indicated two operational taxonomic units, S. agalactiae and S. dysgalactiae (p = 0.009), were present. These results provide a first look at the microbiome as a component of human sperm RNA sequencing that has sufficient sensitivity to identify contamination or potential pathogenic bacterial colonization at least among the known contributors.

Nanopore native RNA sequencing of a human poly(A) transcriptome.

High throughput cDNA sequencing technologies have advanced our understanding of transcriptome complexity and regulation. However, these methods lose information contained in biological RNA because the copied reads are often short and because modifications are not retained. We address these limitations using a native poly(A) RNA sequencing strategy developed by Oxford Nanopore Technologies (ONT). Our study generated 9.9 million aligned sequence reads for the human cell line GM12878, using thirty MinION flow cells at six institutions. These native RNA reads had a median length of 771 bases, and a maximum aligned length of over 21,000 bases. Mitochondrial poly(A) reads provided an internal measure of read length quality. We combined these long nanopore reads with higher accuracy short-reads and annotated GM12878 promoter regions, to identify 33,984 plausible RNA isoforms. We describe strategies for assessing 3′ poly(A) tail length, base modifications, and transcript haplotypes.

Identification of Fusion Transcripts from Unaligned RNA-Seq Reads Using ChimeRScope.

Fusion transcripts that are frequent in cancer can be exploited to understand the mechanisms of malignancy and can serve as diagnostic or prognostic markers. Several algorithms have been developed to predict fusion transcripts from DNA or RNA data. The majority of these algorithms align sequencing reads to the reference transcriptome for predicting fusions; however, this results in several undetected fusions due to the highly perturbed nature of cancer genomes. Here, we describe a novel method that uses a k-mer based algorithm to predict fusion transcripts accurately using the unaligned reads from the regular RNA-seq data analysis pipelines.

Microsatellite instability status is determined by targeted sequencing with MSIcall in 25 cancer types

BackgroundMicrosatellite instability (MSI) occurs in solid tumors and is a predictive biomarker for remarkable response to immune checkpoint inhibitors. Detection of MSI status has been conventionally conducted by PCR–electrophoresis-based assay (MSI-PCR) and immunohistochemistry (IHC) of mismatch repair proteins. However, these approaches require visual confirmation and involve some difficulties in determining MSI statuses from equivocal results. MethodsWe performed amplicon-based targeted sequencing of 76 microsatellite loci (MSI-NGS) in 184 formalin-fixed paraffin-embedded (FFPE) tumor tissues and baseline control samples. A bioinformatics tool, MSIcall, was used to calculate the quantitative values based on the aligned sequence reads and evaluated MSI status. Furthermore, we examined the concordance between the results from MSI-NGS and MSI-PCR/IHC. Diagnostic accuracy, sensitivity, and specificity were estimated by receiver operating characteristic (ROC) curve analysis. For validation cohort, we studied additional 50 tumor samples to determine the MSI status. ResultsOf 184 tumor samples, MSI-PCR and IHC analysis classified 161 tumors as MSS/pMMR and 23 as MSI-H/dMMR. Using MSI-NGS combined with MSIcall, we predicted MSI status with high accuracy (98.9%), specificity (91.3%), and sensitivity (100%) in 25 types of cancers. This method achieved an area under the ROC curve (AUC) value of 0.9986. Furthermore, we achieved the 100% concordant results using additional 50 samples for validation. ConclusionWe demonstrated newly developed MSI-NGS with MSIcall accurately determines the MSI status of FFPE tumor tissues thorough sequencing of tumor samples alone without patient-matched normal controls. This approach can be applied to all types of solid tumors to determine responders to immune-oncology therapy.

Large scale microbiome profiling in the cloud.

MotivationBacterial metagenomics profiling for metagenomic whole sequencing (mWGS) usually starts by aligning sequencing reads to a collection of reference genomes. Current profiling tools are designed to work against a small representative collection of genomes, and do not scale very well to larger reference genome collections. However, large reference genome collections are capable of providing a more complete and accurate profile of the bacterial population in a metagenomics dataset. In this paper, we discuss a scalable, efficient and affordable approach to this problem, bringing big data solutions within the reach of laboratories with modest resources.ResultsWe developed Flint, a metagenomics profiling pipeline that is built on top of the Apache Spark framework, and is designed for fast real-time profiling of metagenomic samples against a large collection of reference genomes. Flint takes advantage of Spark’s built-in parallelism and streaming engine architecture to quickly map reads against a large (170 GB) reference collection of 43 552 bacterial genomes from Ensembl. Flint runs on Amazon’s Elastic MapReduce service, and is able to profile 1 million Illumina paired-end reads against over 40 K genomes on 64 machines in 67 s—an order of magnitude faster than the state of the art, while using a much larger reference collection. Streaming the sequencing reads allows this approach to sustain mapping rates of 55 million reads per hour, at an hourly cluster cost of $8.00 USD, while avoiding the necessity of storing large quantities of intermediate alignments.Availability and implementation Flint is open source software, available under the MIT License (MIT). Source code is available at https://github.com/camilo-v/flint.Supplementary information Supplementary data are available at Bioinformatics online.