Multi-biomarker panel selection on a GPU

Liquid chromatography-based tandem mass spectrometry (LC-MS) technique allows for identification and quantification of thousands of proteins in parallel. This technique coupled with a feed-forward artificial neural network provides a technique to analyze and select protein panels for use in multi-biomarker panel discovery applications. In this study, we enhance this technique by utilizing massively parallel computation enabled by a high-end Graphics Processing Unit (GPU). We utilize a GPU-based back-propagation feed-forward artificial neural network to help select an optimal panel of protein biomarkers for breast cancer diagnosis. By exploiting the GPU particularly for accelerating optimal biomarker panel discovery, we achieved a computation speedup of 32.2X over a comparable sequential program implemented on a CPU. GPUs have become a cost-effective alternative, offering end-user high-performance computing alternative to computer cluster or cloud computing. We showed how to achieve substantial improvement in computation using domain-specific parallel computing on a GPU. This approach can be generalized to other bioinformatics problems.