OBGene: Optimal barcode generator for NGS experiments

We are addressing a fundamental problem in the selection of optimal sets of DNA barcodes to be used in multiplexing of samples in a next-generation sequencing (NGS) through an optimization framework called OBGene. We define a constrained-optimization setting whose cost function and constraints utilize the notion of Hamming distance between barcodes and the experimental requirements, respectively. We consider diverse possible constraints that might be imposed by a practitioner; for example, maximum distance between barcodes, balanced list, and a desired limit for the GC-content. Owing to the size of the search space, we propose to use a stochastic search algorithm, ICA, to find the global optima of the cost function. As a numerical example, we provide a list of 20 barcodes through the simulation.