Biological Computing: A Convergence of Nanotechnology and LifeInspired Computation

Chapter 1: Biological computing

For the purpose of carrying out digital or real computations, biological computers make use of molecules that are derived from biological systems, such as DNA and/or proteins.

Nanobiotechnology, a relatively recent field of study, has proved instrumental in facilitating the development of biocomputers. The term nanobiotechnology can be defined in a number of different ways. In a more general sense, nanobiotechnology can be described as any sort of technology that makes use of both nano-scale materials (i.e. materials that have typical dimensions of 1-100 nanometers) and materials that are based on biological processes. According to a definition that is more limiting, nanobiotechnology is defined as the process of designing and engineering proteins, which may subsequently be combined into bigger structures that have functional properties.

Scientists now have the ability to construct biomolecular systems in such a way that they interact in a manner that can ultimately result in the computing functionality of a computer. This is made possible by the deployment of nanobiotechnology, which is described in this more precise meaning.

In order to carry out computing tasks, biocomputers make use of elements that are produced from biological sources. On the basis of the conditions (input) of the system, a biocomputer is made up of a pathway or series of metabolic pathways that involve biological components. These pathways are created to function in a particular manner. The output, which is based on the engineering design of the biocomputer and can be viewed as a sort of computational analysis, is the pathway of reactions that takes place as a result of the biocomputer's operation. Biochemical computers, biomechanical computers, and bioelectronic computers are the three unique categories of biocomputers that can be distinguished from one another.

The enormous variety of feedback loops that are characteristic of biological chemical reactions is utilized by biochemical computers in order to accomplish the task of achieving computational functionality. The feedback loops that are present in biological systems can take on a variety of forms. Furthermore, numerous elements have the potential to offer both positive and negative feedback to a specific biochemical process. This feedback can result in either an increase or a decrease in the amount of chemical output, depending on the specific situation. Some examples of such factors are the quantity of catalytic enzymes that are present, the quantity of reactants that are present, the quantity of products that are there, and the presence of molecules that bind to any of the aforementioned components and, as a result, affect the chemical reactivity of any of the previous factors. It is possible to engineer a chemical pathway that consists of a set of molecular components that react to produce one particular product under one set of specific chemical conditions and another particular product under another set of conditions. This is possible because of the nature of these biochemical systems, which is to be regulated through a wide variety of different mechanisms. It is possible for the existence of the specific product that is produced as a result of the pathway to function as a signal. This signal, together with other chemical signals, can be interpreted as a computational output based on the initial chemical conditions of the system, which are referred to as the input.

In the same way as biochemical computers and biomechanical computers both carry out a particular operation, which may be regarded as a functional calculation dependent on particular initial conditions that serve as input, biomechanical computers are comparable to biochemical computers. One thing that differentiates them, however, is the specific signal that is used as the output. The output signal of biochemical computers is determined by the presence of particular chemicals or the concentration of particular molecules. The output of biomechanical computers, on the other hand, is the mechanical shape of a particular molecule or set of molecules under a certain set of beginning conditions. Biomechanical computers are dependent on the characteristics of particular molecules, which allow them to take on particular physical configurations when subjected to particular chemical conditions. The product of the biomechanical computer is analyzed in order to determine its mechanical, three-dimensional structure, which is then interpreted in an appropriate manner during the calculation process.

There is also the possibility of constructing biocomputers in order to carry out electrical computing. In the same way that calculations are carried out by biomechanical and biochemical computers, computations are carried out by interpreting a particular output that is based upon an initial set of conditions that act as input. The nature of the electrical conductivity that is detected in the bioelectronic computer is the nature of the output that is measured in bioelectronic computers.  Depending on the initial conditions that serve as the input of the bioelectronic system, this output is comprised of proteins that have been carefully designed to conduct electricity in extremely particular fashions so as to achieve the desired results.

The field of networks-based biocomputation involves the exploration of a microscopic network that encodes a mathematical problem of interest by biological agents that are self-propelled. Examples of such agents include molecular motor proteins and bacteria. There are a number of alternative solutions to the problem, and they can be represented by the pathways that the agents take through the network or by their end positions. As an illustration, movable molecular motor filaments are identified at the exits of a network that encodes the NP-complete problem SUBSET SUM in the system that was developed by Nicolau et al. The right solutions to the algorithm are represented by each and every exit that these filaments visit. Non-solutions are exits that have not been visited. It is either actin and myosin or kinesin and microtubules that are responsible for the movement of the cell. Myosin and kinesin, in that order, are connected to the bottom of the network channels, respectively. The addition of adenosine triphosphate (ATP) causes the actin filaments or microtubules to be driven through the channels, which allows them to investigate the network under investigation. When compared to other forms of computing, such as electrical computing, the conversion of energy from chemical energy (ATP) to mechanical energy (motility) is extremely efficient. As a result, the computer, in addition to being massively parallel, also utilizes orders of magnitude less energy for each step of the computation.

Biologically derived computational systems like these are dependent on the specific molecules that comprise the system. These molecules are usually proteins, but they may also include DNA molecules. The behavior of these systems is dependent on molecules.  The creation of such a system requires the synthesis of a number of different chemical components, which can be accomplished through the use of nanobiotechnology. The order in which amino acids are arranged in a protein is what determines its chemical composition. Amino acids are the fundamental chemical components of proteins. To a certain extent, this sequence is determined by a particular sequence of DNA nucleotides, which are the fundamental components of DNA molecules. Nucleotide sequences are translated into proteins in biological systems by biological molecules known as ribosomes. These ribosomes assemble individual amino acids into polypeptides that form functional proteins based on the nucleotide sequence that the ribosome interprets. Proteins are produced in biological systems. In the end, this indicates that it is possible to engineer the chemical components that are required to produce a biological system that is capable of doing calculations. This is accomplished by engineering the sequences of DNA nucleotides to encode for the essential protein components. Additionally, it is possible for the DNA molecules that have been synthetically produced to exhibit functionality within a specific biocomputer system. In light of this, the application of nanobiotechnology to the design and production of synthetically created proteins, as well as the design and synthesis of artificial DNA molecules, can make it possible to construct functioning biocomputers, such as computational genes.

It is also possible to develop biocomputers with cells serving as their fundamental building blocks.  The production of logic gates from individual cells can be accomplished through the use of chemically induced dimerization systems. The activation of these logic gates is caused by chemical agents that generate contacts between proteins that were not previously interacting with one another and cause a change in the cell that can be readily observed.

Network-based biocomputers are created through the process of nanofabrication, which involves the fabrication of hardware from wafers. The channels are etched using either electron-beam lithography or nano-imprint lithography methods. To ensure that the protein filaments are directed in the desired direction, the channels have been constructed to have a high aspect ratio of cross section. Additionally, split and pass junctions are designed in such a way that filaments will propagate throughout the network and investigate the paths that are permitted.  The process of surface silanization guarantees that the motility proteins will be able to adhere to the surface while still maintaining their functionality. It is the molecules that are derived from biological tissue that are responsible for performing the logic operations.

The capacity to self-replicate and self-assemble into functional components is a characteristic shared by all examples of biological entities. The potential for all biologically derived systems to self-replicate and self-assemble, provided that the conditions are suitable, is the source of the economic gain that biocomputers offer. For example, a single DNA molecule may be used to produce multiple copies of all of the proteins that are required for a certain biochemical route. Furthermore, this pathway could be modified to function as a biocomputer. All of these proteins could be generated within a biological cell.  It would then be possible to reproduce this DNA molecule an infinite number of times. Because of this property of biological molecules, the manufacturing of these molecules could be extraordinarily effective while also being quite inexpensive. Biocomputers, on the other hand, may be manufactured in vast quantities from cultures without the need for any additional technology when it comes to their assembly, in contrast to electronic computers, which require manual production.

Biocomputers are currently available with a wide range of functional capabilities, which include the ability to perform mathematical computations and operations involving binary logic. The first person to propose a biochemical computing scheme was Tom Knight, who works at the Massachusetts Institute of Technology's Artificial Intelligence Laboratory. In this system, protein concentrations are utilized as binary signals, which ultimately help to carry out logical processes.  In a biocomputer chemical route, a signal that is either a 1 or a 0 is indicated by a concentration of a specific biochemical product that is either at or above a specified level. A concentration that is below this level signals the other signal that is still present. By employing this strategy as a computational analysis, biochemical computers are able to carry out logical operations in which the required binary output will only take place under particular logical limitations on the initial conditions. Putting it another way, the suitable binary output functions as a conclusion that is obtained logically from a collection of initial conditions that serve as premises from which the logical conclusion can be made. Additionally, it has been demonstrated that biocomputers are capable of performing other functional capacities, such as mathematical computations, in addition to these types of logical processes already mentioned. One such example was offered by W.L. Ditto, who, in 1999, developed a biocomputer at Georgia Tech including neurons derived from leeches. This biocomputer was able to do addition operations of a straightforward nature. Simply told, these are just a few of the notable applications that biocomputers have already been designed to carry out, and the capabilities of biocomputers are growing increasingly advanced. As was mentioned earlier, the development of biocomputer technology is a popular and fast expanding area of research that is anticipated to see a great deal of progress in the future. This is due to the fact that the production of biomolecules and biocomputers is both readily available and has the potential to be economically efficient.

The term transcriptor was used by a group of bioengineers from Stanford University, led by Drew Endy, who made the announcement in March 2013 that they had developed the biological counterpart of a transistor.  A fundamental logic system, data storage, and information transmission are the three components that are required to construct a fully functional computer. The invention was the third and last component amongst these three components.

Parallel biological computing with networks, in which the movement of bio-agents corresponds to the addition of arithmetic, was shown in 2016 on a SUBSET SUM instance that contained eight candidate solutions.

It was published in Nature in July 2017 that separate investigations conducted with E. coli demonstrated the possibility of utilizing living cells for the purpose of computing tasks and storing information. The Wyss Institute for Biologically Inspired Engineering at Harvard University and the Biodesign Institute at Arizona State University collaborated to form a team that produced a biological computer inside of E. coli that was able to respond to a dozen different inputs. The group of people used the term ribocomputer to refer to the computer because it was made up of ribonucleic acid. After successfully preserving photos and movies in the DNA of living E. coli cells, researchers from Harvard demonstrated that it is possible to preserve information in bacteria after demonstrating that it is possible to do so.

An investigation into the notion of distributed computing among cells was carried out in 2021 by a group of researchers lead by biophysicist Sangram Bagh. The research involved the use of E. coli to solve challenges involving a two-by-two maze.

In the year 2024, FinalSpark, a biocomputing business based in Switzerland, introduced an online platform that made it possible for researchers all over the world to carry out studies on biological neurons in vitro remotely.

There are numerous instances of straightforward biocomputers that have been created; nevertheless, the capabilities of these biocomputers are extremely restricted in comparison to those of non-biocomputers that are available for commercial use.

The potential to solve complex mathematical problems using a significantly lower amount of energy than standard electronic supercomputers, as well as the ability to perform more reliable calculations simultaneously rather than sequentially, is what motivates the continued development of scalable biological computers, and a number of funding agencies are supporting these efforts.

{End Chapter 1}

Chapter 2: Reactome

Reactome is a database of biological pathways that may be accessed online for free. Manual curation and authorship of the content is performed by biologists with a doctoral degree, in conjunction with the editorial staff of Reactome. Cross-references have been made between the content and a variety of bioinformatics databases. Reactome was developed with the intention of providing a visual representation of biological pathways that include all of the mechanical details, while also making the source material available in a format that is accessible to computational systems.

An international multidisciplinary team from the Office of the International Center for Research Resources (OICR), Ohio State University (OHSU), the European Molecular Biology Laboratory (EMBL-EBI), and New York University Medical Center (NYULMC) is responsible for maintaining Reactome. This team possesses expertise in pathway curation and annotation, software development, as well as training and outreach. Their goal is to provide the research community with openly accessible biological pathway knowledge. Lincoln Stein, who is the OICR, is the leader of the Reactome team. Guanming Wu from Ohio State University, Peter D'Eustachio from New York University Medical Center, and Henning Hermjakob from EMBL-EBI. Through the use of email, the Reactome helpdesk can be contacted. 

via the usage of the website, one is able to navigate via paths and send data to a collection of data analysis tools. Several standard formats, such as PDF, SBML, Neo4j GraphDB, MySQL, PSI-MITAB, and BioPAX, are available for download in their entirety. The data that lies behind the surface is also available. Pathway diagrams are constructed using a style that is based on the Systems Biology Graphical Notation (SBGN) and Systems Description (PD).

The response serves as the fundamental building block of the Reactome data model. Nucleic acids, proteins, complexes, and tiny molecules are the entities that are involved in reactions. These entities come together to form a network of biological interactions, which are then categorized into pathways. Signal transmission, innate and acquired immunological function, transcriptional control, programmed cell death, and classical intermediate metabolism are some examples of biological processes that can be found in the Reactome.

The routes that are represented in Reactome are distinct to each species, and each step of the pathway is backed up by literature citations that include an experimental verification of the process that is being represented. It is possible for routes to contain steps that have been manually inferred from non-human experimental details if there is no experimental verification using human reagents. However, this is only possible if an expert biologist, who is referred to as the Author of the pathway, and another biologist, who is referred to as the Reviewer, are in agreement that this is a valid inference to make. In order to computationally build routes in other organisms that are derived from the human pathway, an orthology-based technique is