Article
Biochemical Research Methods
Yoo-Jin Ha, Seungseok Kang, Jisoo Kim, Junhan Kim, Se-Young Jo, Sangwoo Kim
Summary: This study benchmarks 11 approaches for detecting mosaic single-nucleotide variant and insertion-deletion mutations.
Article
Biochemical Research Methods
Jeffrey N. Dudley, Celine S. Hong, Marwan A. Hawari, Jasmine Shwetar, Julie C. Sapp, Justin Lack, Henoke Shiferaw, Jennifer J. Johnston, Leslie G. Biesecker
Summary: PBVI is a method for Position-Based Variant Identification that improves the detection of low-level mosaic variants by using empirically-derived distributions of alternate nucleotides from a control dataset. Modeled on 11 segmental overgrowth genes, this method shows superior detection of single nucleotide mosaic variants of 0.01-0.05 variant allele fraction. Sensitivity of over 85% and 95% was observed at depths of 600 x and 1200 x, respectively, demonstrating its utility in identifying pathogenic variants in individuals with somatic overgrowth disorders.
BMC BIOINFORMATICS
(2021)
Article
Biology
Dries Decap, Louise de Schaetzen van Brienen, Maarten Larmuseau, Pascal Costanza, Charlotte Herzeel, Roel Wuyts, Kathleen Marchal, Jan Fostier
Summary: Halvade Somatic is a somatic variant calling pipeline that leverages Big Data processing platforms, enabling the processing of large volumes of sequencing data in a short amount of time with reliable and scalable performance.
Article
Multidisciplinary Sciences
Ramesh Rajaby, Dong-Xu Liu, Chun Hang Au, Yuen-Ting Cheung, Amy Yuet Ting Lau, Qing-Yong Yang, Wing-Kin Sung
Summary: This article introduces a new method called INSurVeyor for detecting insertions and compares its sensitivity with other methods. It also provides updated catalogues of insertions in Arabidopsis and human genomes based on INSurVeyor.
NATURE COMMUNICATIONS
(2023)
Article
Biochemical Research Methods
Sunho Lee, Seokchol Hong, Jonathan Woo, Jae-hak Lee, Kyunghee Kim, Lucia Kim, Kunsoo Park, Jongsun Jung
Summary: A new method called RDscan has been developed to improve the accuracy of germline and somatic variant calling in NGS data by calculating RDscore and removing false-positive variant calls. Testing showed that RDscan significantly improved accuracy for most algorithms, particularly in enhancing the accuracy of somatic variants.
JOURNAL OF COMPUTATIONAL BIOLOGY
(2022)
Review
Biochemistry & Molecular Biology
Monica Valecha, David Posada
Summary: Single-cell sequencing data is error-prone due to technical biases, and there are various methods for single-cell variant calling that often result in many discordant calls when applied to real data.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Article
Oncology
R. Tyler McLaughlin, Maansi Asthana, Marc Di Meo, Michele Ceccarelli, Howard J. Jacob, David L. Masica
Summary: Accurately identifying somatic mutations is crucial for precision oncology and calculating tumor-mutational burden (TMB), which predicts response to immunotherapy. This study applies machine learning models to classify somatic vs germline mutations in tumor-only solid tumor samples, achieving state-of-the-art performance. The addition of a machine-learning classifier improves the concordance of TMB estimates and eliminates racial bias in tumor-only variant calling.
NPJ PRECISION ONCOLOGY
(2023)
Article
Biology
Lanying Wei, Martin Dugas, Sarah Sandmann
Summary: SimFFPE is a read simulator designed for FFPE samples, while FilterFFPE is a filtration algorithm that improves the positive predictive value of structural variants and maintains sensitivity.
Article
Biochemical Research Methods
Nae-Chyun Chen, Alexey Kolesnikov, Sidharth Goel, Taedong Yun, Pi-Chuan Chang, Andrew Carroll
Summary: In this study, population-aware DeepVariant models were developed to improve the accuracy and recall of variant calling in single samples. By using allele frequencies from the 1000 Genomes Project, this model reduced variant calling errors and improved the precision of rare homozygous and pathogenic clinvar calls. The study also found that diverse reference panels were more accurate than population-specific panels, even when the sample ancestry matched the population.
BMC BIOINFORMATICS
(2023)
Article
Biochemical Research Methods
Carlos A. Garcia-Prieto, Francisco Martinez-Jimenez, Alfonso Valencia, Eduard Porta-Pardo
Summary: This study highlights the significant impact of variant calling decisions on key analyses of cancer genomic data, such as the identification of cancer driver genes, quantification of mutational signatures, and detection of clinically actionable variants. The choice of somatic variant caller or strategy to combine them can lead to important differences in results, potentially impacting treatment decisions for patients. The widely used Consensus of three calling strategy may result in the loss of important cancer driver genes and actionable mutations, emphasizing the importance of careful consideration in variant calling for critical analyses of tumor sequencing data.
Article
Biochemical Research Methods
Sergey Vilov, Matthias Heinig
Summary: In this study, a convolutional neural network-based approach called DeepSom is presented for detecting somatic single nucleotide polymorphism and short insertion and deletion variants in tumor whole-genome sequencing samples without a matched normal. The performance of DeepSom is validated on five different cancer datasets and it is demonstrated to outperform previously proposed methods for tumor-only somatic variant calling on whole-genome sequencing samples.
Article
Biology
Tanveer Ahmad, Zaid Al Ars, H. Peter Hofstee
Summary: The article proposes a high-performance workflow for variant calling methods based on deep learning and introduces efficient data transfer between different workflow stages using Python and Apache Arrow. The design outperforms existing implementations by over 2 times in preprocessing stages, creating a scalable and high-performance solution for DeepVariant.
Article
Biotechnology & Applied Microbiology
Daniel P. Cooke, David C. Wedge, Gerton Lunter
Summary: Octopus is a variant caller that uses a polymorphic Bayesian genotyping model capable of modeling different experimental designs within a unified haplotype-aware framework. It accurately calls germline variants in individuals, including low-frequency somatic variations, while producing fewer false positives compared to other methods. Octopus also outputs realigned evidence BAM files to assist with validation and interpretation.
NATURE BIOTECHNOLOGY
(2021)
Article
Multidisciplinary Sciences
Jindrich Adolf, Yoram Segal, Matyas Turna, Tereza Novakova, Jaromir Dolezal, Patrik Kutilek, Jan Hejda, Ofer Hadar, Lenka Lhotska
Summary: The objective of this study is to evaluate the performance of functional tests using a camera-based system and machine learning techniques. The results suggest that a camera-based system combined with machine learning algorithms can be a simple and inexpensive tool to assess the performance quality of functional tests.
Article
Biotechnology & Applied Microbiology
Desiree Schnidrig, Andrea Garofoli, Andrej Benjak, Gunnar Ratsch, Mark A. Rubin, Salvatore Piscuoglio, Charlotte K. Y. Ng
Summary: Precision oncology relies on accurately identifying somatic mutations in cancer patients. PipeIT2, a variant calling workflow, addresses the clinical need to define somatic mutations in the absence of germline control and achieves high recall and reliability in mutation identification.