MGI Tech Co., Ltd

MegaBOLT - Bioinformatics Analysis Accelerator

MegaBOLT bioinformatics analysis accelerator series (including MegaBOLT/ZBOLT/ZBOLT Pro, hereinafter called MegaBOLT) is an MGI self-developed and MPS-concentrated hardware accelerating system for bioinformatics analysis. ZBOLT Pro/ZBOLT/MegaBOLT supports the analysis of Whole Genome Sequencing (WGS), Whole Exome Sequencing (WES), and Panel Sequencing on Germline or Somatic data. It is up to 300 times faster than the classical algorithm. Fast, integrated, easy to use, and economic are MegaBOLT`s most prominent features.MegaBOLT products series include the workstation (MegaBOLT) and rack server (ZBOLT/ZBOLT Pro), which can be applied to various scenarios of all scales.

Most popular related searches

bioinformatics analysis

bioinformatics

genomic sequencing

vertebral compression fracture

genomics

bioinformatics platform

MegaBOLT adopts heterogeneous computing to accelerate analysis. It accelerates algorithms such as SOAPnuke, Minimap2, BWA, GATK HaplotypeCaller + MuTect2 through multi-stream and highly parallelized computing architecture, which contributes to a massive speedup in massive data analysis.

* Basic: Includes the analysis steps from clean data after quality control to variant calls.
* Full: Includes extra steps of quality control and report generating in addition to the Basic workflow.
* The time consumed is measured by taking the average of completing 4 identical tasks (MegaBOLT) / 12 (ZBOLT) / 48 (ZBOLT Pro).

Multi-task scheduling system acceleration
Integrated with our self-developed multi-task scheduling system, MegaBOLT supports simultaneous multi-task computing on a single server. The computing efficiency is further improved by 15%~31%.

01. Precision
After speedup, the analysis precision remains steadfast. MegaBOLT is also configured with a deep learning variant calling module, MegaBOLT-DV, which is optimized by algorithms and specified neural network model training. When MegaBOLT is used in combination with PCR-Free library preparation and MGI DNBSEQ™ sequencing technologies, the human WGS variant calling precision can reach SNP 99.9% and INDEL 99%.

02. Deep learning making sequencing more cost-effective
The deep learning variant calling module, MegaBOLT-DV, helps to greatly reduce the sequencing depth and cost without affecting the analysis precision.

Compared with the traditional variant calling results of 30X PCR data, SNP of 15X PCR-Free data from MegaBOLT-DV is close while INDEL is better.

Integrated
Integrated with the self-developed laboratory information management system ZLIMS, MegaBOLT realizes completely automated sequencing and analysis. It can work together with MGI DNBSEQ™ sequencers or sequencers of other platforms.

Laboratory device management and real-time monitoring
Sample information recording and tracking
Data transmission and management
Analysis task scheduling and processing
Analysis result/report management and visualization

Easy to use
The Web design makes MegaBOLT simple and easy to use. You can just click a couple of buttons to submit a task and check the results on the website. In this case, a bioinformatic technical background is no longer necessary.

Economic
With a prominent speedup in computing and rigid control over hardware prime cost, a MegaBOLT server works like 10 normal computers in total and is without doubt highly cost-effective.

Heterogeneous computing accommodates high-speed sequence analysis
MegaBOLT applies heterogeneous computing to the workflow from FASTQ input of sequencing data to BAM output of alignment result and VCF/gVCF output of variant calling result.

We also further streamlined the whole process and developed a highly parallelized computing architecture, thus realizing highspeed WGS/WES analysis and computing.

Compared with the common analysis workflow, MegaBOLT stands out for the following features:

Specialized hardware logic circuit for Bioinformatics computing
Multi-stream computing architecture
Highly parallelized computing architecture