GIInger predicts homologous recombination deficiency and patient response to PARPi treatment from shallow genomic profiles.

Homologous recombination deficiency (HRD) is a predictive biomarker for poly(ADP-ribose) polymerase 1 inhibitor (PARPi) sensitivity. Routine HRD testing relies on identifying BRCA mutations, but additional HRD-positive patients can be identified by measuring genomic instability (GI), a consequence of HRD. However, the cost and complexity of available solutions hamper GI testing. We introduce a deep learning framework, GIInger, that identifies GI from HRD-induced scarring observed in low-pass whole-genome sequencing data. GIInger seamlessly integrates into standard BRCA testing workflows and yields reproducible results concordant with a reference method in a multisite study of 327 ovarian cancer samples. Applied to a BRCA wild-type enriched subgroup of 195 PAOLA-1 clinical trial patients, GIInger identified HRD-positive patients who experienced significantly extended progression-free survival when treated with PARPi. GIInger is, therefore, a cost-effective and easy-to-implement method for accurately stratifying patients with ovarian cancer for first-line PARPi treatment.

Benefits of applying molecular barcoding systems are not uniform across different genomic applications.

BACKGROUND: Despite the wide variety of Next Generation Sequencing (NGS)-based methods, it remains challenging to detect mutations present at very low frequencies. This problem is particularly relevant in oncology, where the limiting amount of input material, and its low quality, often limit the performance of the assays. Unique Molecular Identifiers (UMIs) are a molecular barcoding system often coupled with computational methods of noise suppression to improve the reliability of detection of rare variants. Although widely adopted, UMI inclusion imposes additional technical complexity and sequencing cost. Currently, there are no guidelines on UMI usage nor a comprehensive evaluation of their advantage across different applications. METHODS: We used DNA sequencing data generated by molecular barcoding and hybridization-based enrichment, from various types and quantities of input material (fresh frozen, formaldehyde-treated and cell-free DNA), to evaluate the performance of variant calling in different clinically relevant contexts. RESULTS: Noise suppression achieved by read grouping based on fragment mapping positions ensures reliable variant calling for many experimental designs even without exogenous UMIs. Exogenous barcodes significantly improve performance only when mapping position collisions occur, which is common in cell-free DNA. CONCLUSIONS: We demonstrate that UMI usage is not universally beneficial across experimental designs and that it is worthwhile to critically consider the comparative advantage of UMI usage for a given NGS application prior to experimental design.

Improved NGS-based detection of microsatellite instability using tumor-only data.

Microsatellite instability (MSI) is a molecular signature of mismatch repair deficiency (dMMR), a predictive marker of immune checkpoint inhibitor therapy response. Despite its recognized pan-cancer value, most methods only support detection of this signature in colorectal cancer. In addition to the tissue-specific differences that impact the sensitivity of MSI detection in other tissues, the performance of most methods is also affected by patient ethnicity, tumor content, and other sample-specific properties. These limitations are particularly important when only tumor samples are available and restrict the performance and adoption of MSI testing. Here we introduce MSIdetect, a novel solution for NGS-based MSI detection. MSIdetect models the impact of indel burden and tumor content on read coverage at a set of homopolymer regions that we found are minimally impacted by sample-specific factors. We validated MSIdetect in 139 Formalin-Fixed Paraffin-Embedded (FFPE) clinical samples from colorectal and endometrial cancer as well as other more challenging tumor types, such as glioma or sebaceous adenoma or carcinoma. Based on analysis of these samples, MSIdetect displays 100% specificity and 96.3% sensitivity. Limit of detection analysis supports that MSIdetect is sensitive even in samples with relatively low tumor content and limited microsatellite instability. Finally, the results obtained using MSIdetect in tumor-only data correlate well (R=0.988) with what is obtained using tumor-normal matched pairs, demonstrating that the solution addresses the challenges posed by MSI detection from tumor-only data. The accuracy of MSI detection by MSIdetect in different cancer types coupled with the flexibility afforded by NGS-based testing will support the adoption of MSI testing in the clinical setting and increase the number of patients identified that are likely to benefit from immune checkpoint inhibitor therapy.

High-Throughput Nucleotide Resolution Predictions of Assay Limitations Increase the Reliability and Concordance of Clinical Tests.

PURPOSE: The ability of next-generation sequencing (NGS) assays to interrogate thousands of genomic loci has revolutionized genetic testing. However, translation to the clinic is impeded by false-negative results that pose a risk to patients. In response, regulatory bodies are calling for reliability measures to be reported alongside NGS results. Existing methods to estimate reliability do not account for sample- and position-specific variability, which can be significant. Here, we report an approach that computes reliability metrics for every genomic position and sample interrogated by an NGS assay. METHODS: Our approach predicts the limit of detection (LOD), the lowest reliably detectable variant fraction, by taking technical factors into account. We initially explored how LOD is affected by input material amount, library conversion rate, sequencing coverage, and sequencing error rate. This revealed that LOD depends heavily on genomic context and sample properties. Using these insights, we developed a computational approach to predict LOD on the basis of a biophysical model of the NGS workflow. We focused on targeted assays for cell-free DNA, but, in principle, this approach applies to any NGS assay. RESULTS: We validated our approach by showing that it accurately predicts LOD and distinguishes reliable from unreliable results when screening 580 lung cancer samples for actionable mutations. Compared with a standard variant calling workflow, our approach avoided most false negatives and improved interassay concordance from 94% to 99%. CONCLUSION: Our approach, which we name LAVA (LOD-aware variant analysis), reports the LOD for every position and sample interrogated by an NGS assay. This enables reliable results to be identified and improves the transparency and safety of genetic tests.

The tobacco genome sequence and its comparison with those of tomato and potato.

The allotetraploid plant Nicotiana tabacum (common tobacco) is a major crop species and a model organism, for which only very fragmented genomic sequences are currently available. Here we report high-quality draft genomes for three main tobacco varieties. These genomes show both the low divergence of tobacco from its ancestors and microsynteny with other Solanaceae species. We identify over 90,000 gene models and determine the ancestral origin of tobacco mosaic virus and potyvirus disease resistance in tobacco. We anticipate that the draft genomes will strengthen the use of N. tabacum as a versatile model organism for functional genomics and biotechnology applications.