Year: 2022
Type: preprint
Abstract: Studying the heterogeneity in cancerous tissue is challenging in cancer research. It is vital to process the real-world data efficiently to understand the heterogeneous nature of cancer tissue. GPU compatible models, which can estimate the subpopulation of cancerous tissue, are fast if the size of input data, i.e., the number of qPCR (quantitative polymerase chain reaction) gene expression reading is extensive. In the real world, we rarely get that much data to reap the benefits of a GPU’s parallelism. Real experimental data from fibroblasts are much less, and models using those data on a GPU are slower than the CPU multithreaded application. This paper will show a method to run GPU-compatible models for cancer tissue heterogeneity on a multithreaded CPU. Further, we also show that the model running on a multithreaded CPU is faster than the model running on a GPU with real experimental data.
Institution: Georgia Institute of Technology
Language: English
Subfield: Hardware and Architecture
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4295073737
