Vecuum

Vecuum identifies vector-originated sequencing reads and false variant calls to detect and remove vector contamination that biases somatic variant detection at low variant allele frequencies.

Key Features:

• Exon-junction clipping analysis: Analyzes clipping patterns at exon junctions to identify reads originating from vectors.
• Intron-less cDNA discrimination: Exploits the intron-less cDNA structure of typical vector inserts, which produces distinct clipping characteristics versus genuine sample sequences.
• False-variant detection via allele-bias: Detects false variant calls by examining biased distributions of mutant alleles associated with vector-originated reads.
• Validation performance: Validated on simulated and spike-in experimental data with a 93% detection rate for vector contaminants and removal of up to 87% of false variant calls at 100% precision.
• Public-dataset screening: Applied to public sequencing datasets to identify false variants resulting from external contamination.
• Implementation: Implemented in Java.

Scientific Applications:

• Vector contamination screening: Identification of vector-originated reads in sequencing datasets for contamination control.
• Somatic variant call refinement: Reduction of false-positive somatic variant calls at low variant allele frequencies.
• Data quality control: Detection of contamination-derived false variants in both experimental spike-ins and public sequencing datasets.

Methodology:

Vecuum analyzes clipping patterns at exon junctions to detect reads derived from intron-less cDNA vector inserts and examines biased distributions of mutant alleles among these reads to flag false variant calls.