Accurate calling of low-frequency somatic mutations by sample-specific modeling of error rates DOI Creative Commons

Yixin Lin,

Carmen Oroperv,

Peter Sørud Porsgård

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 20, 2024

Abstract Calling rare somatic variants from NGS data is more challenging than calling inherited variants, especially if the variant only present in a small fraction of cells sequenced biopsy. In this case, having good estimate error rate specific base particular read becomes essential. paired-end sequencing, where some DNA fragments are shorter twice length, overlapping regions pairs an ideal resource for training models to discern context-dependent rates, as any discordant bases overlaps must be caused by sequencing or alignment error. We have created new tool named BBQ (an acronym Better Base Quality) that uses reads conditional on mutation type, sequence context, and quality. also how much concordant decreased compared non-overlapping reads. Results show can remove errors induced damage increased quality differs between samples types, reflecting different patterns samples. use call variants. Sequencing testis biopsy cell-free sample serve proof-of-concept germ cell detecting cancer mutations. find using sample-specific allows us with fewer false positives existing tools such Mutect2 Strelka2.

Language: Английский

Beyond basics: Key mutation selection features for successful tumor‐informed ctDNA detection DOI Creative Commons
Marijana Nesic, Mads H. Rasmussen, Tenna Vesterman Henriksen

et al.

International Journal of Cancer, Journal Year: 2024, Volume and Issue: 155(5), P. 925 - 933

Published: April 16, 2024

Tumor-informed mutation-based approaches are frequently used for detection of circulating tumor DNA (ctDNA). Not all mutations make equally effective ctDNA markers. The objective was to explore if prioritizing using mutational features-such as cancer cell fraction (CCF), multiplicity, and error rate-would improve the success rate tumor-informed analysis. Additionally, we aimed develop a practical easily implementable analysis pipeline identifying candidate from whole-exome sequencing (WES) data. We analyzed WES data three studies, one on bladder (Cohort A) two colorectal (Cohorts I N). studies included 390 patients. For each patient, unique set (median mutations/patient: 6, interquartile 13, range: 1-46, total n = 4023) were markers ctDNA. tool PureCN assess CCF multiplicity mutation. High-CCF detected more than low-CCF A: odds ratio [OR] 20.6, 95% confidence interval [CI] 5.72-173, p 1.73e-12; Cohort I: OR 2.24, CI 1.44-3.52, 1.66e-04; N: 1.78, 1.14-2.79, 7.86e-03). detection-likelihood additionally improved by selecting with or above 1.55, 1. 14-2.11, 3.85e-03; 1.23-2.56, 1.34e-03; 1.94, 1.63-2.31, 2.83e-14). Furthermore, which method had lowest rates, detection-likelihood, particularly evident when plasma cell-free fractions below 0.1% (p 2.1e-07). Selecting high CCF, low significantly likelihood. provide free access enabling others perform qualified prioritization

Language: Английский

Citations

4
The role of renal and liver function in clinical ctDNA testing DOI Creative Commons
John A. Conrad, Tenna Vesterman Henriksen, Jesper Nors

et al.

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(2), P. e0319194 - e0319194

Published: Feb. 25, 2025

Circulating tumor DNA (ctDNA) has high clinical potential in early cancer detection. The renal system and the liver are involved clearing circulating cell free (cfDNA) from blood. Recent studies on mice show that inhibiting liver’s ability to clear cfDNA results elevated ctDNA levels blood samples. Emphasizing need for humans exploring if markers of function associated with present study investigates level, level detection is affected colorectal (CRC) patients biomarkers indicative low function. We requisitioned standard laboratory tests function, measured within thirty days curative intended surgery 846 stage I-III CRC patients. For each patient, matching preoperative data was available. investigated correlation between impaired findings revealed variation did not affect or test remain stable over a wide range biomarker results.

Language: Английский

Citations

0
Detection of circulating tumor DNA by tumor-informed whole-genome sequencing enables prediction of recurrence in stage III colorectal cancer patients DOI Creative Commons
Amanda Frydendahl, Jesper Nors,

Mads H. Rasmussen

et al.

European Journal of Cancer, Journal Year: 2024, Volume and Issue: 211, P. 114314 - 114314

Published: Sept. 11, 2024

Language: Английский

Citations

3
Evaluating Bioinformatics Processing of Somatic Variant Detection in cfDNA Using Targeted Sequencing with UMIs DOI Open Access

Yixin Lin,

Mads H. Rasmussen, Mikkel Christensen

et al.

International Journal of Molecular Sciences, Journal Year: 2024, Volume and Issue: 25(21), P. 11439 - 11439

Published: Oct. 24, 2024

Circulating tumor DNA (ctDNA) is a promising cancer biomarker, but accurately detecting mutations in cell-free (cfDNA) challenging due to their low frequency and sequencing errors. Our study benchmarked Mutect2, VarScan2, shearwater, DREAMS-vc using deep targeted of cfDNA with Unique Molecular Identifiers (UMIs) from 111 colorectal patients. Performance was assessed at both the mutation level (distinguish variants errors) sample (detect if an individual has cancer). Additionally, we investigated effects various UMI grouping consensus strategies. The shearwater-AND variant calling method demonstrated highest precision tumor-derived plasma, reached ROC-AUC 0.984 for classification tumor-informed analyses. exhibited 0.808 tumor-agnostic studies. We also found that depth differences PBMCs could lead false positives, particularly VarScan2 which addressed by downsampling equivalent mean depths. network-based methods outperformed those identical UMIs when all reads were retained. findings emphasize optimal caller depends on context—whether focused or classification, whether conducted under conditions.

Language: Английский

Citations

1
A panorama of colon cancer in the era of liquid biopsy DOI Creative Commons
Sylvie Devalle, Veronica Aran,

Cesar de Souza Bastos Júnior

et al.

The Journal of Liquid Biopsy, Journal Year: 2024, Volume and Issue: 4, P. 100148 - 100148

Published: March 13, 2024

Colon cancer (CC) is one of the most frequent cancers worldwide being responsible for over 500 thousand deaths in 2022. Its financial and human burden expected to increase next decades accompanying growing aging global population. Much this could be alleviated considering that lethality CC mostly due its late diagnosis failure individualized management patients. Coordinated government actions implementation better diagnostic tools capable detecting earlier tracking tumoral evolution are mandatory achieve a reduction CC's social impact. CtDNA-based liquid biopsy (LB) has great potential contribute patients' screening adhesion, detection, longitudinal tumor follow-up. In review, we will discuss latest epidemiological data on disease, diagnostic, subtypes, genetics, treatment focusing advantages limitations ctDNA-based LB, including important bottlenecks solutions necessary clinical translation. The ctDNA-directed trials also examined.

Language: Английский

Citations

0
Stool and Blood DNA Tests for Colorectal Cancer Screening DOI
Thomas F. Imperiale, Kyle Porter, Paul J. Limburg

et al.

New England Journal of Medicine, Journal Year: 2024, Volume and Issue: 390(23), P. 2222 - 2225

Published: June 19, 2024

Language: Английский

Citations

0
Accurate calling of low-frequency somatic mutations by sample-specific modeling of error rates DOI Creative Commons

Yixin Lin,

Carmen Oroperv,

Peter Sørud Porsgård

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 20, 2024

Abstract Calling rare somatic variants from NGS data is more challenging than calling inherited variants, especially if the variant only present in a small fraction of cells sequenced biopsy. In this case, having good estimate error rate specific base particular read becomes essential. paired-end sequencing, where some DNA fragments are shorter twice length, overlapping regions pairs an ideal resource for training models to discern context-dependent rates, as any discordant bases overlaps must be caused by sequencing or alignment error. We have created new tool named BBQ (an acronym Better Base Quality) that uses reads conditional on mutation type, sequence context, and quality. also how much concordant decreased compared non-overlapping reads. Results show can remove errors induced damage increased quality differs between samples types, reflecting different patterns samples. use call variants. Sequencing testis biopsy cell-free sample serve proof-of-concept germ cell detecting cancer mutations. find using sample-specific allows us with fewer false positives existing tools such Mutect2 Strelka2.

Language: Английский

Citations

0