Bioinfomatics
DaBoost “parallel multi-pattern” search technology speeds up software and hardware processes for DNA Sequencing.
DaBoost are looking for forward thinking Bioinformatics companies to develop strategic partnerships.
Email us
up to
entries for DNA patterns
from
base pairs
up to
base pairs
Fast full gene and mutation detection across all genomes can be in excess of 10 times faster, reducing the time to prognosis dramatically. NGS processes can also run significantly quicker with “parallel multi-pattern” search technology and can reduce the need for high-end computation.
- For each pattern multiple uniqueness keys are selected automatically to improve statistics
- Key selection is using typical data information pre-prepared using the human genome to achieve uniqueness
- Preferred key selection is using the basic ACGT Nucleic but upon using other letters, multiple entries are defined
- Pattern dictionary insertion using ‘Regular’ and ‘Reversed’ Nucleic order
- Performing check for ‘Insertion’, ‘Deletion’ and ‘Replacement’ problems on patterns
The basic technology updates “one-by-one pattern scan” or “tree symbols decision” searches into “parallel multi-pattern” searches to speed up digital processes by up to 100x using faster pattern search engine based on unique smart & highly efficient dictionary technology and can handle both text and binary.
After fast identifying the file hierarchy and file sections using this technology it is much easier to parse the required sections or parallelize the sections parsing. We can make significant impacts to binary as well as text and have identified four initial major market vectors including AI/Big data, Cyber Security, DNA/NGS Sequencing and Hardware Acceleration.
This technology can be used inside communication protocols, file or stream parsing, DPI – deep packet inspection filtering and protocol stacks. Hardware modifications can increase speed even more. We currently can identify file modes but will soon be able to manage streaming and block modes.
The ongoing development of basic libraries with broad applications will be delivered along with the tools to easily integrate custom functionality. Both software and hardware companies will be able to license the technology for further development and customization
We have a working Linux lite version library (97kb program size = 32byte search, max 100 thousand parallel searches) and the full version (200kb program size = 2Gigabyte search, max 100 million parallel searches) out in Q2 and will support Windows and Mac.
A new and exciting development is also underway to run the program in a web browser which we believe will be a game changer for us due later in 2022.
