FPGA Acceleration of a hierarchical clustering and data fusion application for disease subtype discovery

Starting from an R implementation of the hierarchical clustering and data fusion algorithm, the method is translated into C++ and further optimized via software techniques for a better execution time and scalability with the number of patients and number of omics. Once the software optimizations reach saturation in terms of time improvement, the code is used to design hardware accelerators which allow parallel execution and other hardware specific optimizations using an FPGA as a targeted device.