A machine for calculating genetic ability estimates

PC-8850BA customer who saw our machine for gene expression analysis (September 2022 version) inquired about a machine for performing calculations using genomic data and statistical analysis.
SNP (snip/single nucleotide polymorphism) genotype information is used to perform calculations to estimate genetic ability. I hear you need more memory.

In addition, the machine specs currently in use are as follows.

・The model is Mac Pro
・CPU: Xeon E5-2697v2 2.7GHz 12 cores
・ Memory: 64GB

The program I'm using is BLUPF90, and I know that 64GB memory is not enough to handle multiple calculations.

As for other uses, I hear that in the future there are plans to use Python for data analysis of 16S ribosomal RNA and machine learning using livestock big data.

The OS is Windows, which you are familiar with, and the storage capacity must be 4 TB or more.
The customer commented, "I'm not familiar with PCs, so it's difficult to make a specific request, but I would like you to propose a machine with as high a spec as possible within the budget."

Based on these considerations, we propose the following specifications.

【Main Specifications】

CPU AMD Ryzen ThreadripperPRO 5975WX (3.60GHz 32 cores)
memory 256GB
Storage 1 4TB SSD S-ATA
Storage 2 16TB HDD S-ATA
video NVIDIA Geforce RTX3090
network on board (1GbE x1 10GBase-T x1)
Housing + power supply Tower type housing + 1000W
OS Microsoft Windows 11 Pro 64 bit
Others TPM module

Based on the case study No. PC-8850B you saw, we proposed a modified configuration.

■ Points

・Considering the main use, it is thought that the processing in the CPU will be the main.
・Since the memory is 64GB and it is "slightly insufficient", it is speculated that if there is more capacity, the same processing can be executed with a margin.

For the CPU, we chose the ThreadripperPRO 5975WX, which has a well-balanced number of cores and clocks.
Also, the memory is 256 GB with a margin.This is equivalent to four times the capacity of the machine currently in use, so it is assumed that the same processing can be carried out with relative leeway.

In addition, considering the possibility of incorporating machine learning methods in the future, it is equipped with a high-end GPU that is often used for machine learning.Since there are various types of machine learning implementations,If you don't use the GPGPU framework that supports NVIDIA GPUs or programs via the CUDA Toolkit, or if you don't intend to work on it in earnest, it doesn't necessarily have to be this GPU.So you can keep the price down by changing to a cheaper product.

In addition, compared to the Mac Pro you are using, there is a large generation difference in the parts, so it is a configuration that allows you to experience the improvement in performance.

The configuration of this case study is based on the conditions given by the customer.
Please feel free to contact us even if you are considering different conditions from what is posted.


・What is the BLUPF90 series program?
The BLUEPF90 family of programs is a statistical software package for mixture model calculations in animal husbandry.
Used in quantitative genetics approaches for animal and plant breeding.
A program written in Fortran 90/95 can perform genomic selection from hundreds of thousands of genotypes.

reference:BLUPF90 Family of Programs *Jumps to an external site


・What is machine learning?
Machine learning is a mechanism that allows computers to perform specific tasks by accumulating data.
Computers autonomously improve recognition and prediction accuracy.
For machine learning, please see related articles on our owned media "TEGAKARI".

reference:[Special article] What is machine learning?