(1) Understand bioinformatics
Bioinformatics uses the methods of applied mathematics, informatics, statistics and computer science to study biological problems. The research materials and results of bioinformatics are all kinds of biological data. The research tools are computers, and the research methods include searching (collecting and screening), processing (editing, organizing, managing and displaying) and Utilize (computation, simulation).
Typical workflow in bioinformatics
This process consists of a series of chained steps that transform raw input (sequencing raw data RAW) into meaningful or interpretable output, for example, fastq files generated from high-throughput sequencing NGS data. Specific tools for specific functional aspects of genome sequence analysis are then executed. Depending on the type of analysis performed, a workflow can have a variable number of steps and thus be simple or complex.
The main research directions of bioinformatics: DNA/RNA/protein sequencing, sequence comparison, gene discovery, genome assembly, drug design, drug discovery, protein structure comparison and prediction, by using computationally intensive techniques (for pattern recognition, data mining, machine learning algorithms, and visualization) to deepen understanding of biological processes. Therefore, it is necessary to be equipped with advanced computing equipment and rich professional analysis software
(2) Computational features of bioinformatics analysis
Many people may have thought about how to choose an ideal graphics workstation hardware configuration:
What is the best PC/workstation for bioinformatics and computational biology research?
Server configuration for bioinformatics analysis
Hardware Configuration of Whole Genome Sequence Analysis Laboratory
What are the requirements for a high-throughput sequence analysis server?
Recommended hardware configuration for next-generation sequencing data analysis
Computer configuration for analysis of NGS metagenomics data?
2.1 Computational features of bioinformatics analysis
Bioinformatics data analysis involves genomics, transcriptomics, proteomics, metagenomics, metabolomics, etc. The following figure shows the whole genome data analysis process
The calculations involved in bioinformatics data analysis are mainly
(1) Sequence/map alignment calculation (Mapping) during resequencing
For mapped reads using programs such as BWA/Bowtie, the memory RAM requirement is not high (for example, 32GB is enough), but the number of CPU cores (and their frequency) will determine how long the calculation process takes. If you are doing a lot of alignment and alignment (e.g. using BWA), then having a lot of CPU cores is more important than having a lot of memory.
Of course configuration specifications depend on your budget and the type of analysis you plan to conduct.
The most computationally intensive step in RNASeq is the comparison step, and the comparison often only needs to be done once! Generally, the hardware configuration of 32-core CPU+64GB RAM can meet the standard mapping and downstream analysis of genome/transcriptome/rainbow genome analysis.
(2) De novo sequence assembly calculation (Assembly)
If you want to perform de novo assembly (such as Velvet), assuming a person's whole genome sequencing data, using the next-generation sequencing method, the human genome is 3G, and the 10-fold data is 30G, then these 30G bases are cut into smaller kmer , assuming that the data has increased to 100G, and some other information of the stored sequence is not counted, all the data must be stored in the memory at one time when the sequence is spliced. If the memory does not reach 100G, the splicing cannot be completed at all.
Therefore, for large-scale genome assembly, a lot of hardware resources are required. It doesn’t matter if the CPU has enough computing power and the memory is more than 150G. But for bacterial genomes, the data set and genome data are not too large, and 128GB of memory is enough.
In order to maximize the workload of NGS (Next Generation Sequencing) analysis, there are three key bottlenecks in hardware configuration: available CPU cores, memory capacity, and I/O bandwidth
2.2 Bioinformatics analysis requires hardware configuration
How to handle 454 and Illumina data? Whole genome assembly/assembly? Sequence splicing? Mapping reads to a reference genome?
(1) How much storage space is needed to keep data read in real time (hard disk capacity)?
The bottleneck in developing clinical applications of next-generation sequencing (high-throughput sequencing) is the storage and analysis of the large amounts of data generated. The applications are diverse, but the common theme is computationally and analytically challenging.
(2) How big is each file to be analyzed (RAM capacity, hard disk read and write speed)?
(3) Is there a requirement for the software to be used to be ready to use multiprocessor runtime (number of CPU cores)?
Configuration reference:
(1) Based on the size of the genome project
(2) Based on the number of researchers in the research group
(3) Recommended Graphics Workstation Configuration for Bioinformatics Analysis 2023
(4) Bioinformatics analysis multi-computer cluster configuration recommendation 2023
If you want to inquire about the processing speed of the machine, technical consultation, and request for detailed technical solutions, please contact
