Fig. 1: Workflow of FVC.
From: FVC as an adaptive and accurate method for filtering variants from popular NGS analysis pipelines
Taking the VCF and BAM files as input, FVC uses the feature construction module to build three types of features related to sequence content, sequencing experiment, and bioinformatic analysis process. If a pre-trained model is already built for the specific pipeline, the variants can be immediately classified as true or false using the filtering module. Otherwise, the pre-trained model can be built using the data construction and supervised learning modules of FVC.
