Fujitsu Technology to Solve Combinatorial Optimization Problems for Medium-Sized Drug Discovery

Figure 1: Relationship in between the scale of problems and variety of bits.
Figure 1: Relationship in between the scale of problems and variety of bits.

FujitsuLaboratoriesLtd today revealed that it has actually established a massive issue division technology, produced together with Fujitsu Advanced Technologies Limited, that uses the Digital Annealer, Fujitsu’s computational architecture that can solve combinatorial optimization problems at spectacular speed.

Previously, the Digital Annealer might just manage problems up to a size that might be input into standalone hardware. This brand-new advancement now makes it possible to extract vital parts of an issue based upon its attributes, and after processing with the Digital Annealer, are consistently cycled through to return to a general state. During this phase, the approach to generate the most enhanced action is embraced, that makes it possible to use to massive problems.

By using this technology to the 2nd generation Digital Annealer, which will can dealing with 8 Kbits- scale (8,192 bits) problems, Fujitsu Laboratories expects that it will be possible to use it to problems on a 100 Kbit scale. In this trial, Fujitsu Laboratories validated through simulations that it would be possible to solve the 30 Kbit steady molecular setup search problems anticipated to be utilized in medium-sized drug discovery. This technology can cut simulation time to simply a couple of days from the half a year taken by standard computer systems. Fujitsu Laboratories will start a joint research study with ProteinQure Inc.( 1 ) in order to speed up the discovery of mid-size protein rehabs.

FujitsuLaboratories is contributing to the production of brand-new service by utilizing the brand-new technology and using the Digital Annealer to combinatorial optimization problems in a range of fields.


Combinatorial optimization problems exist throughout a series of fields in daily society, requiring the mixes that will produce the most effective and efficient outcomes. With combinatorial optimization problems, as the variety of mixes increases explosively when all the aspects that require to be thought about to discover services are factored in, there is a limitation to the scale of an issue that can be fixed in an useful length of time with existing computingtechnology In light of this, Fujitsu Laboratories established the Digital Annealer, a devoted architecture for combinatorial optimization problems. Services for 1 Kbit (1,024 bits) scale problems are currently offered, and research study is continuous to offer assistance of up to 8 Kbits, within financial 2018.

There is need, nevertheless, to go even more and offer assistance for enormous problems of a scale higher that those that can be input into hardware so as to solve combinatorial optimization problems of evermore intricacy. For example, with calculations of around 30 Kbits, it ends up being possible to rapidly solve steady setup search problems for particles of up to 50 amino acids in the field of medium-sized drug discovery. This can speed up the advancement of medium-sized drugs. Beyond that, this technology can likewise be used to a range of other fields, consisting of factories and producing in addition to transport.


Simply segmenting massive problems into parts that are little adequate to be input into the hardware, then enhancing those private parts would not enhance the issue as a whole. In other words, while it is possible to enhance private parts by drawing out parts of an issue, if the proper part is not drawn out, an enough level of efficiency can not be accomplished.

About the Newly Developed Technology

FujitsuLaboratories has actually now established a technology to manage problems of a bigger scale than can be input into standalone hardware. When this technology is used to the 2nd generation Digital Annealer, it is anticipated to can dealing with problems up to 100 Kbits in size. The functions of the recently established technology are as follows.

1. Executes an option search cycle in order to discover the ideal service

After carrying out a fast total search of the whole issue, this technology draws out a part of the issue of a size that can be input into the hardware, and utilizes the Digital Annealer to search for an ideal service for the drawn out part. That result is fed back into the issue as an entire, and the cycle is duplicated several times, altering the drawn out part, and discovering a much more ideal service for the larger-scale issue.

2. Multiple division approaches based upon the attributes of the issue

To increase the performance of enhancing the issue as an entire, it is necessary to identify the part to extract, depending upon the attributes of the concern. Fujitsu Laboratories concentrated on the relationship of the issue, establishing several division approaches, consisting of a technique that mainly draws out the parts of the issue as a whole that alter the most quickly, and a technique that sections the parts that have the least quantity of coupling in between aspects. By picking the proper division approach for each issue, this technology can effectively discover services for massive problems.


FujitsuLaboratories, together with ProteinQure Inc., a business establishing a software application platform for protein drug discovery, validated the possibility of utilizing the Digital Annealer on simulations that discover steady setups of protein drug prospects.

In protein drug discovery, drug prospects, where anywhere from a handful to about 50 amino acids are linked in a chain, have an impact as drugs by binding highly with targeted proteins. The Digital Annealer is utilized to search for the most steady setups when each amino acid is designed and put on a point in a 3D grid, utilizing aspects such as the binding relationship in between amino acids. Then the strength of the binding in between the setup found with this approach and the target protein is examined utilizing a docking calculation. By duplicating this cycle about 1,000 times, for example, this technology can discover an extremely efficient medium-sized drug prospect.

By using this technology to the 2nd generation Digital Annealer, Fujitsu Laboratories was able to reduce the simulation time needed for a calculation including a medium-sized drug prospect including 48 amino acids (about 30 Kbits) from numerous hours utilizing standard computer systems to numerous minutes, utilizing the very same approach of modeling amino acids. This suggests that by duplicating this circulation, a search for healing particles that would have taken half a year might be carried out in a couple of days. Applying this recently established technology, the Digital Annealer is anticipated to speed up the advancement of protein rehabs, which are drawing in attention as next-generation drugs.

Medium-Sized Drug DiscoveryFigure 2: Example application of the brand-new technology to a steady setup search issue


By utilizing the Digital Annealer and the application of this technology, in the future Fujitsu Laboratories anticipates to be able to manage enormous combinatorial optimization problems of up to 100,000 bits, contributing to consumers in a large range of companies, consisting of drug discovery, chemistry, production, transport, financing, and logistics.

Accordingly,Fujitsu Laboratories will start a joint research study with ProteinQureInc in order to speed up the commercialization of this technology for protein drug discovery.

Source: FujitsuLaboratories Ltd

Recommended For You

About the Author: livescience

Leave a Reply

Your email address will not be published. Required fields are marked *