Benchmark of GPU-accelerated bioinformatics methods for processing raw RNA-seq data
Description
The emergence of personalized medicine requires being able to produce and process huge amounts of biological data generated from a patients' biological samples, in a quick manner and at a reasonable cost.
While modern sequencing technologies have keep up with these need, and are now able to produce large amount of data in record time, bioinformatics tools still have to make this transformation.
Indeed, most bioinformatics methods focus more on the accuracy of their results than on the speed of their execution. This creates a situation where bioinformatics analysis can create a bootlneck. To remedy that problem, we have to look at ways to speed up the analysis. NVIDIA, one of the world’s largest GPU manufacturer recently released the 3rd version of it’s Clara Parabricks suite, which accelerates populars bioinformatics tools by allowing them to use GPU for theirs calculations. However, Parabricks has only been independently benchmarked on it’s ability to handle genomic data and not RNAseq data. We thus propose to benchmark parabricks on RNAseq data.
Files
Bardet_JOBIM2022.pdf
Files
(636.4 kB)
Name | Size | Download all |
---|---|---|
md5:886f7b7eb5decac407470dee42c18811
|
636.4 kB | Preview Download |