Fig. 1
From: Genomic variant benchmark: if you cannot measure it, you cannot improve it
Importance of benchmark datasets in genomics and clinical setting. These datasets are beneficial in assessing different tools and methods. First, a benchmark dataset is established by using a myriad of dependable variant calling methods, forming a collection of reliable variant calls, normally stored as a VCF file (green box; step 1). New DNA sequencing technologies could be evaluated by comparing their corresponding called variants with the benchmark dataset on the same DNA sample. This showcases the reliability of their sequenced reads. The same process can be used to test new algorithms developed for read alignment or variant calling (step 2). Clinical research studies also benefit from such benchmark datasets, by incorporating newly well-established variant calling methods and sequencing technologies (step 3)