Occurrence of extra-chromosomal circular DNA is a phenomenon frequently observed in tumor cells, and the presence of such DNA has been recognized as a marker of adverse outcome across cancer types. We here describe a computational workflow for identification of DNA circles from long-read sequencing data. The workflow is implemented based on the Snakemake workflow management system. Its key step uses a graph-theoretic approach to identify putative circular fragments validated on simulated reads. We then demonstrate robustness of our approach using nanopore sequencing of selectively enriched circular DNA by highly sensitive and specific recovery of plasmids and the mitochondrial genome, which is the only circular DNA in normal human cells. Finally, we show that the workflow facilitates detection of larger circular DNA fragments containing extrachromosomal copies of the MYCN oncogene and the respective breakpoints, which is a potentially useful application in disease monitoring of several cancer types.

Data from sequencing of DNA or RNA samples is routinely scanned for variation. Such variation data is stored in the standardized VCF/BCF format with additional annotations. Analyses of variants usually involve steps where filters are applied to narrow down the list of candidates for further analysis. A number of tools for this task exist, differing in functionality, speed, syntax and supported annotations. Thus, users have to switch between tools depending on the filtering task, and have to adapt to the respective filtering syntax. We present vembrane as a command line VCF/BCF filtering tool that consolidates and extends the filtering functionality of previous software to meet any imaginable filtering use case. To this end, vembrane exposes the VCF/BCF file type specification and its inofficial extensions by the annotation tools VEP and SnpEff as Python data structures. vembrane filter enables filtration by arbitrary Python expressions over (combinations of) annotations, requiring only basic knowledge of the Python programming language. vembrane table allows users to generate tables from subsets of annotations or functions thereof. Finally, it is fast, thanks to pysam, a Python wrapper around htslib, and by relying on Python's lazy evaluation.

No abstract

Summary We present vembrane as a command line VCF/BCF filtering tool that consolidates and extends the filtering functionality of previous software to meet any imaginable filtering use case. Vembrane exposes the VCF/BCF file type specification and its inofficial extensions by the annotation tools VEP and SnpEff as Python data structures. vembrane filter enables filtration by Python expressions, requiring only basic knowledge of the Python programming language. vembrane table allows users to generate tables from subsets of annotations or functions thereof. Finally, it is fast, by using pysam and relying on lazy evaluation. Availability and Implementation Source code and installation instructions are available at github.com/vembrane/vembrane, DOI:10.5281/zenodo.7003981. Supplementary information Supplementary data are available at Bioinformatics online.