Dataset Information

Comparison of different semi-automated cfDNA extraction methods in combination with UMI-based targeted sequencing.

ABSTRACT: The analysis of circulating cell-free DNA (cfDNA) extracted from peripheral blood can serve as a minimally invasive alternative to tumor tissue biopsies in cases with impaired access to tissue. Its clinical utility has been well demonstrated for EGFR T790M testing in lung cancer patients suffering progress after tyrosine kinase inhibitor treatment. At present, highly sensitive unique molecular identifiers (UMI)-based NGS for liquid biopsy testing is less established compared to single gene assays. However, the critical bottleneck are sufficient cfDNA yields, which are essentially required to obtain meaningful test results. We compared four different cfDNA extraction methods (Qiagen, Promega, Thermo and Stratec) using the same plasma samples in order to evaluate their suitability for further NGS analysis. We managed to draw 60 ml blood from 12 patients each and equally collected 30ml in PAXgene and EDTA tubes at the same time point, sufficient for total of 96 cfDNA extractions. CfDNA concentrations and total amounts were highest for Qiagen and Promega protocols, showing the best read length profiles after sequencing. Known oncogenic driver mutations were identified in 9 out of 12 patients with at least one of the cfDNA extraction methods, again favoring the extraction protocols from Qiagen and Promega. We also uncovered putative sequencing artefacts including known driver genes pointing to a careful consideration for the limit of detection of this methodology. Our study shows that pre-analytical optimization is necessary to achieve the maximum sensitivity of UMI-based sequencing but also highlights the low abundance of tumor-derived cfDNA in lung cancer samples.

SUBMITTER: Streubel A

PROVIDER: S-EPMC6779285 | biostudies-literature | 2019 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Comparison of different semi-automated cfDNA extraction methods in combination with UMI-based targeted sequencing.

Streubel Anna A Stenzinger Albrecht A Stephan-Falkenau Susann S Kollmeier Jens J Misch Daniel D Blum Torsten Gerriet TG Bauer Torsten T Landt Olfert O Am Ende Alexander A Schirmacher Peter P Mairinger Thomas T Endris Volker V

Oncotarget 20191001 55

The analysis of circulating cell-free DNA (cfDNA) extracted from peripheral blood can serve as a minimally invasive alternative to tumor tissue biopsies in cases with impaired access to tissue. Its clinical utility has been well demonstrated for EGFR T790M testing in lung cancer patients suffering progress after tyrosine kinase inhibitor treatment. At present, highly sensitive unique molecular identifiers (UMI)-based NGS for liquid biopsy testing is less established compared to single gene assay ...[more]

PMID: 31620244

Similar Datasets

Project description:BackgroundAutomated ICD coding on medical texts via machine learning has been a hot topic. Related studies from medical field heavily relies on conventional bag-of-words (BoW) as the feature extraction method, and do not commonly use more complicated methods, such as word2vec (W2V) and large pretrained models like BERT. This study aimed at uncovering the most effective feature extraction methods for coding models by comparing BoW, W2V and BERT variants.MethodsWe experimented with a Chinese dataset from Fuwai Hospital, which contains 6947 records and 1532 unique ICD codes, and a public Spanish dataset, which contains 1000 records and 2557 unique ICD codes. We designed coding tasks with different code frequency thresholds (denoted as [Formula: see text]), with a lower threshold indicating a more complex task. Using traditional classifiers, we compared BoW, W2V and BERT variants on accomplishing these coding tasks.ResultsWhen [Formula: see text] was equal to or greater than 140 for Fuwai dataset, and 60 for the Spanish dataset, the BERT variants with the whole network fine-tuned was the best method, leading to a Micro-F1 of 93.9% for Fuwai data when [Formula: see text], and a Micro-F1 of 85.41% for the Spanish dataset when [Formula: see text]. When [Formula: see text] fell below 140 for Fuwai dataset, and 60 for the Spanish dataset, BoW turned out to be the best, leading to a Micro-F1 of 83% for Fuwai dataset when [Formula: see text], and a Micro-F1 of 39.1% for the Spanish dataset when [Formula: see text]. Our experiments also showed that both the BERT variants and BoW possessed good interpretability, which is important for medical applications of coding models.ConclusionsThis study shed light on building promising machine learning models for automated ICD coding by revealing the most effective feature extraction methods. Concretely, our results indicated that fine-tuning the whole network of the BERT variants was the optimal method for tasks covering only frequent codes, especially codes that represented unspecified diseases, while BoW was the best for tasks involving both frequent and infrequent codes. The frequency threshold where the best-performing method varied differed between different datasets due to factors like language and codeset.

Project description:BACKGROUND: Different genome annotation services have been developed in recent years and widely used. However, the functional annotation results from different services are often not the same and a scheme to obtain consensus functional annotations by integrating different results is in demand. RESULTS: This article presents a semi-automated scheme that is capable of comparing functional annotations from different sources and consequently obtaining a consensus genome functional annotation result. In this study, we used four automated annotation services to annotate a newly sequenced genome--Arcobacter butzleri ED-1. Our scheme is divided into annotation comparison and annotation determination sections. In the functional annotation comparison section, we employed gene synonym lists to tackle term difference problems. Multiple techniques from information retrieval were used to preprocess the functional annotations. Based on the functional annotation comparison results, we designed a decision tree to obtain a consensus functional annotation result. Experimental results show that our approach can greatly reduce the workload of manual comparison by automatically comparing 87% of the functional annotations. In addition, it automatically determined 87% of the functional annotations, leaving only 13% of the genes for manual curation. We applied this approach across six phylogenetically different genomes in order to assess the performance consistency. The results showed that our scheme is able to automatically perform, on average, 73% and 86% of the annotation comparison and determination tasks, respectively. CONCLUSIONS: We propose a semi-automatic and effective scheme to compare and determine genome functional annotations. It greatly reduces the manual work required in genome functional annotation. As this scheme does not require any specific biological knowledge, it is readily applicable for genome annotation comparison and genome re-annotation projects.

Dataset Information

Comparison of different semi-automated cfDNA extraction methods in combination with UMI-based targeted sequencing.

Publications

Comparison of different semi-automated cfDNA extraction methods in combination with UMI-based targeted sequencing.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets